ADSORPTION, DESORPTION AND STABILIZATION OF...
Transcript of ADSORPTION, DESORPTION AND STABILIZATION OF...
ADSORPTION, DESORPTION, AND STABILIZATION OF
ARSENIC ON ALUMINUM SUBSTITUTED FERRIHYDRITE
A Thesis
by
YOKO MASUE
Submitted to the Office of Graduate Studies of
Texas A&M University in partial fulfillment of the requirements for the degree of
MASTER OF SCIENCE
December 2004
Major Subject: Soil Science
ADSORPTION, DESORPTION, AND STABILIZATION OF
ARSENIC ON ALUMINUM SUBSTITUTED FERRIHYDRITE
A Thesis
by
YOKO MASUE
Submitted to Texas A&M University in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE
Approved as to the style and content by:
Richard H. Loeppert (Chair)
Timothy A. Kramer (Member)
Charles T. Hallmark
(Member) Mark Hussey
(Head of Department)
December 2004
Major Subject: Soil Science
iii
ABSTRACT
Adsorption, Desorption, and Stabilization of Arsenic
on Aluminum Substituted Ferrihydrite.
(December 2004)
Yoko Masue, B.S., Texas A&M University
Chair of Advisory Committee: Dr. Richard H. Loeppert
Because of As toxicity, the complexity of its chemistry, and the recent lowering
of the maximum contaminant level of As in municipal drinking water, there has been
considerable interest for improved methods to remove As from water. Although Al and
Fe hydroxides have been extensively studied as adsorbents for As removal during water
treatment, coprecipitated Al:Fe hydroxides have received only minimal attention. The
theoretical and experimental feasibility of coprecipitated Al:Fe hydroxide systems were
evaluated by studying their mineralogy, stability, and As adsorption and desorption
behavior.
The broad XRD peaks revealed that Al was substituted into the ferrihydrite
structure and that this was the only major product up to about a 2:8 Al:Fe molar ratio.
Gibbsite and bayerite were identified when Al content was higher. The rate of
recrystallization of ferrihydrite into goethite and hematite was significantly reduced as Al
substitution was increased.
In general, adsorption capacity of both AsV and AsIII decreased with increase in
Al:Fe molar ratio; however, similar AsV adsorption capacities were observed with Fe and
iv
Al:Fe hydroxides with Al:(Al+Fe) molar ratios < 0.20. Both AsIII and AsV adsorption
isotherms were effectively described by Langmuir and Freundlich equations. Adsorption
maxima of AsV on Fe and Al:Fe hydroxides were observed at pH 3 to 7, and that of AsV
on Al hydroxide was observed at pH 5.2, with significant decreases in adsorption with
increase and decrease in pH. Adsorption maxima of AsIII decreased by approximately 4
% for each 10 % increase in Al substitution up to 5:5 Al:Fe molar ratio. Adsorption
maxima of AsIII on Fe and Al:Fe hydroxides were observed at pH 8 to 9. AsIII adsorption
on Al hydroxide was negligible. Counterion Ca2+, compared to Na+, enhanced the
retention of AsV, especially at pH > 7. Counterion concentration did not significantly
affect AsV adsorption. Though phosphate desorbed both AsV and AsIII from all Al:Fe
hydroxides, quantitative desorption was never observed.
The results of this study indicate the possible utility of coprecipitated Al:Fe
hydroxide in wastewater treatment. Based on adsorption/desorption behavior and
stability of the Al:Fe hydroxide product, the preferred Al:Fe molar ratio was 2:8.
v
ACKNOWLEDGEMENTS
I wish to express my gratitude to all of the people who supported me through this
study. I would like to thank Dr. R. H. Loeppert, chair of my advisory committee, for his
enduring encouragement and support. I would especially like to express my appreciation
to Dr. C. T. Hallmark for his support as my undergraduate advisor, soil judging coach,
and a committee member. I would also like to thank my committee member Dr. Timothy
Kramer for his time and assistance in understanding water treatment from the engineering
point of view. I also thank Dr. B. K. Biswas for analytical support, and Dr. G. N. White
and Dr. J. B. Dixon for assistance with mineral characterization and TEM study
respectively. I also thank Lea Dell Morris for her assistance in obtaining library
resources. I also acknowledge the help of my colleagues Brandon Lafferty and Sujin
Yean. Thank you also to Joe Stobaugh and his family for their continuous support and
encouragement. Most of all, I would like to thank my family in Japan for their enduring
love. Without each of these people, this research would not have been possible.
vi
TABLE OF CONTENTS
Page
ABSTRACT....................................................................................................................... iii
ACKNOWLEDGEMENTS................................................................................................ v
TABLE OF CONTENTS................................................................................................... vi
LIST OF TABLES........................................................................................................... viii
LIST OF FIGURES ........................................................................................................... ix
INTRODUCTION .............................................................................................................. 1
OBJECTIVES..................................................................................................................... 3
LITERATURE REVIEW ................................................................................................... 4
Arsenic in the Environment .............................................................................................4 Chemical Properties of Arsenic .......................................................................................4 Arsenic Toxicity ..............................................................................................................9 Critical Problems ...........................................................................................................10 Bonding Mechanisms of Arsenic...................................................................................10 Macroscopic Adsorption Behavior of Arsenic ..............................................................12 Counterion Effect on Arsenic Adsorption .....................................................................13 Effect of Ionic Strength on Arsenic Adsorption ............................................................15 Adsorption Modeling.....................................................................................................16 Mineralogy of Mixed Al:Fe Hydroxides .......................................................................18 Arsenic Removal by Coagulation of Al and Fe Salts ....................................................19 Desorption of Arsenic by Phosphate .............................................................................20 Arsenic Analysis by Flow-Injection Hydride-Generation Flame-Atomic-Absorption Spectroscopy................................................................................................................. 21
MATERIALS AND METHODS...................................................................................... 23
Chemicals ......................................................................................................................23 Synthesis of Ferrihydrite and Al-substituted Analogs...................................................23 Characterization of Synthesized Ferrihydrite and Al-substituted Analogs ...................24
X-ray Diffraction .......................................................................................................24 Transmission Electron Microscopy ...........................................................................24 Point of Zero Salt Effect ............................................................................................24
Stability of Ferrihydrite and Al-substituted Analogs.....................................................25 Analysis .........................................................................................................................27 Arsenic Adsorption Isotherms .......................................................................................27 Arsenic Adsorption Modeling .......................................................................................28
Langmuir Adsorption Isotherms................................................................................28
vii
Page
Freundlich Adsorption Isotherms ..............................................................................28 Arsenic Adsorption Envelopes ......................................................................................28 Arsenic Adsorption as Affected by Counterion.............................................................29 Arsenic Desorption Envelopes ......................................................................................30
RESULTS AND DISCUSSION....................................................................................... 32
Characterization of Synthesized Ferrihydrite and Al-substituted Analogs ...................32 X-ray Diffraction .......................................................................................................32 Transmission Electron Microscopy ...........................................................................39 Point of Zero Salt Effect ............................................................................................46
Stability of Ferrihydrite and Its Al-substituted Analogs................................................51 Effect of Al Substitution and pH ...............................................................................51 Kinetics of Mineral Transformation ..........................................................................54 Effect of Arsenic........................................................................................................59
Arsenic Adsorption Isotherms .......................................................................................59 Effect of Mineralogy..................................................................................................59 Effect of pH ...............................................................................................................70
Arsenic Adsorption Models...........................................................................................71 Langmuir Adsorption Model .....................................................................................71 Freundlich Adsorption Model....................................................................................85 The Langmuir Model versus the Freundlich Model ..................................................98
Arsenic Adsorption Envelopes ......................................................................................98 Adsorption of Arsenic as Affected by Counterion ......................................................104
Adsorption Isotherms of AsV as Affected by Counterion........................................104 Adsorption Envelopes of AsV as Affected by Counterion .......................................107 Hypotheses...............................................................................................................122
Desorption Envelopes of Arsenic ................................................................................124
CONCLUSIONS............................................................................................................. 133
Mineralogy and Stability of the Hydroxides as Affected by Al:Fe Molar Ratio.........133 Adsorption of AsV and AsIII as Affected by Al:Fe Molar Ratio..................................133 Desorption of AsV and AsIII as Affected by Al:Fe Molar Ratio ..................................134 Implications to the Water Treatment ...........................................................................135
REFERENCES ............................................................................................................... 137
VITA............................................................................................................................... 145
viii
LIST OF TABLES
Page
Table 1. Selected reduction half-reactions and thermodynamic constants (Sparks, 2003). ................................................................................................... 6
Table 2. The pKa values of AsIII (H3AsO3) and AsV (H3AsO4) (Wagman et al., 1982)... 6
Table 3. Ca arsenate hydrate precipitates (Bothe and Brown, 1999).............................. 14
Table 4. Percent of ammonium oxalate extractable (AOE) Fe to total Fe following incubation at 70 oC for 96 h. ............................................................................ 51
Table 5. Calculated adsorption maxima (b), binding constants (KL), and linear Langmuir functions of three sets of hypothetical data........................... 72
Table 6. Coefficients of determination (r2), calculated adsorption maxima (b), and binding constants (KL) of AsV adsorption as affected by Al:Fe molar ratios derived by Langmuir linear functions. ............................................................. 80
Table 7. Coefficients of determination (r2), calculated adsorption maxima (b), and binding constants (KL) of AsIII adsorption as affected by Al:Fe molar ratios derived by Langmuir linear functions. ............................................................. 80
Table 8. Calculated N and KF of Freundlich linear functions of two sets of hypothetical data. .................................................................................................................. 86
Table 9. Coefficients of determination (r2), and calculated N and KF values for AsV adsorption by Al:Fe hydroxides, as derived using the Freundlich linear functions........................................................................................................... 92
Table 10. Coefficients of determination (r2), and calculated N and KF values for AsIII adsorption by Al:Fe hydroxides, as derived using the Freundlich linear functions........................................................................................................... 92
Table 11. Proportion of AsV adsorbed during the 24 h adsorption reaction before phosphate desorption (A), proportion of adsorbed AsV after 24 h shaking with deionized water (DIW) (B), and proportion of AsV desorbed during 24 h shaking with deionized water......................................................................... 125
Table 12. Proportion of AsIII adsorbed during the 24 h adsorption reaction before phosphate desorption (A), proportion of adsorbed AsIII after 24 h shaking with deionized water (DIW) (B), and proportion of AsIII desorbed during 24 h shaking with deionized water......................................................................... 125
ix
LIST OF FIGURES
Page
Figure 1. Speciation of AsIII as a function of pH. .............................................................. 7
Figure 2. Speciation of AsV as a function of pH................................................................ 8
Figure 3. Schematic illustration of the bidentate binuclear surface structure of AsV on Fe hydroxide. ....................................................................................... 11
Figure 4. Experimental procedure for determination of stability of the 0:1 and 2:8 Al:Fe hydroxides....................................................................................... 26
Figure 5. XRD patterns of freshly prepared hydroxides at various Al:Fe molar ratios... 33
Figure 6. XRD patterns of fresh 3:7 Al:Fe hydroxides and 3:7 Al:Fe hydroxides aged at 2ºC for 1 month. .................................................................................. 35
Figure 7. XRD patterns of 1:0 and 5:5 Al:Fe hydroxides................................................ 36
Figure 8. Structure of gibbsite. ........................................................................................ 38
Figure 9. Structure of bayerite. ........................................................................................ 38
Figure 10. Colors of freeze dried ferrihydrite and its Al-substituted analogs. ................. 38
Figure 11. TEM micrographs of the 0:1 Al:Fe hydroxides. (a) Aggregates of 0:1 Al:Fe hydroxide. (b) Dense aggregates of 0:1 Al:Fe hydroxide.................... 40
Figure 12. TEM micrographs of the 0:1 Al: Fe hydroxides. (a) Aggregates of 0:1 Al:Fe hydroxide (magnified picture of Figure 11-a). (b) Dense aggregates of 0:1 Al:Fe hydroxide (magnified picture of Figure 11-b)............................. 41
Figure 13. TEM micrographs of the 2:8 Al:Fe hydroxides. (a) Dense aggregates of the 2:8 Al:Fe hydroxide and porous surrounding. (b) Dense aggregates of 2:8 Al:Fe hydroxide with dispersed aggregates........................................... 42
Figure 14. TEM micrographs of the 2:8 Al:Fe hydroxides. (a) Aggregates of 2:8 Al:Fe hydroxide (magnified from Figure 13-a). (b) Aggregates of 2:8 Al:Fe hydroxide (magnified from Figure 13-b). ........................................................ 43
Figure 15. TEM micrographs of the 5:5 Al:Fe hydroxides. (a) Gibbsite, bayerite, and small aggregates of the 5:5 Al:Fe hydroxide. (b) Magnified image of bayerite crystal found in the 5:5 Al:Fe hydroxide. .......................................... 44
x
Page
Figure 16. TEM micrographs of the 5:5 and 0:1 Al:Fe hydroxides. (a). Magnified image of gibbsite crystal found in the 5:5 Al:Fe hydroxide. (b) Gibbsite, bayerite, and aggregate of Al hydroxide found in the 0:1 Al:Fe hydroxide. ... 45
Figure 17. Titration curves for the 0:1 Al:Fe hydroxide at three ionic strengths. ............ 47
Figure 18. Titration curves for the 2:8 Al:Fe hydroxide at three ionic strengths. ............ 48
Figure 19. Titration curves for the 5:5 Al:Fe hydroxide at three ionic strengths. ............ 49
Figure 20. Titration curves for the 1:0 Al:Fe hydroxide at three ionic strengths. ............ 50
Figure 21. XRD patterns of the hydroxides incubated without AsV at 70 oC for 96 h. .... 53
Figure 22. The activities of hydrolysis species of Fe3+ in equilibrium with amorphous Fe hydroxide as a function of pH, calculated using thermodynamic constants from Lindsay (1979). ............................................. 55
Figure 23. The influence of incubation time on the proportion of added AsV not adsorbed. The incubation was at pH 10 and 70 oC. The AsV (0.05 mmolAs mmolAl+Fe
-1) was added to the suspension following incubation and was allowed to equilibrate at pH 7 and 23 ºC for 2 h.............. 56
Figure 24. The influence of incubation time on the proportion of added AsV not adsorbed. The incubation was at pH 4 and 70 oC. The AsV (0.05 mmolAs mmolAl+Fe
-1) was added to the suspension following incubation and was allowed to equilibrate at pH 7 and 23 ºC for 2 h.............................................. 57
Figure 25. AsV remaining in solution from the final incubation products as a function of proportion of crystalline Fe not extracted by pH 3 ammonium oxalate from the products. The AsV was added to the suspension following incubation and was allowed to equilibrate at pH 7 and 23 ºC for 2 h.............. 58
Figure 26. Adsorption isotherms of AsV at pH 5 on precipitated products of various Al:Fe molar ratio. ................................................................................ 60
Figure 27. Adsorption isotherms of AsV at pH 8 on precipitated products of various Al:Fe molar ratio. ................................................................................ 61
Figure 28. Adsorption isotherms of AsV at low equilibrium AsV concentrations at pH 5 on precipitated products of various Al:Fe molar ratio. ....................... 63
Figure 29. Adsorption isotherms of AsV at low equilibrium AsV concentrations at pH 8 on precipitated products of various Al:Fe molar ratio. ....................... 64
xi
Page
Figure 30. Adsorption isotherms of AsIII at pH 5 on precipitated products of various Al:Fe molar ratio. ................................................................................ 66
Figure 31. Adsorption isotherms of AsIII at pH 8 on precipitated products of various Al:Fe molar ratio. ............................................................................................. 67
Figure 32. Adsorption isotherms of AsIII at low equilibrium AsIII concentrations at pH 5 on precipitated products of various Al:Fe molar ratio. ....................... 68
Figure 33. Adsorption isotherms of AsIII at low equilibrium AsIII concentrations at pH 8 on precipitated products of various Al:Fe molar ratio. ....................... 69
Figure 34. Langmuir adsorption isotherms for three sets of hypothetical data with fixed adsorption maxima (b) and varying bonding strength (KL) as summarized in Table 5. .................................................................................... 73
Figure 35. Linear regression using the linear form of the Langmuir equation for three sets of hypothetical data with fixed adsorption maxima (b) and varying bond strength (KL), as summarized in Table 5. .................................. 74
Figure 36. Langmuir adsorption isotherms of three sets of hypothetical data with fixed intercept of the linear Langmuir equation (1/KLb) and increasing adsorption maxima (b), as summarized in Table 5. ......................................... 75
Figure 37. Linear regression using the linear form of Langmuir equation for three sets of hypothetical data with fixed intercept (1/KLb) and increasing adsorption maxima (b), as summarized in Table 5. ......................................... 76
Figure 38. Langmuir adsorption isotherms of three sets of hypothetical data with fixed binding strength (KL) and varying adsorption maxima (b), as summarized in Table 5. .................................................................................... 77
Figure 39. Linear regression using the linear form of the Langmuir equation for three sets of hypothetical data with fixed binding strength (KL) and varying adsorption maxima (b), as summarized in Table 5. ............................ 78
Figure 40. Linear regression using the linear form of the Langmuir equation for evaluation of adsorption of AsV by Al:Fe hydroxides at pH 5, as summarized in Table 6. .................................................................................... 81
Figure 41. Linear regression using the linear form of the Langmuir equation for evaluation of adsorption of AsV by Al:Fe hydroxides at pH 8, as summarized in Table 6. .................................................................................... 82
xii
Page
Figure 42. Linear regression using the linear form of the Langmuir equation for evaluation of adsorption of AsIII by Al:Fe hydroxides at pH 5, as summarized in Table 7. .................................................................................... 83
Figure 43. Linear regression using the linear form of the Langmuir equation for evaluation of adsorption of AsIII by Al:Fe hydroxides at pH 8, as summarized in Table 7. .................................................................................... 84
Figure 44. Freundlich adsorption isotherms, q = KFCN, for three sets of hypothetical data, with fixed KF but with varying N, as summarized in Table 8. ................ 87
Figure 45. Linear regression lines of the Freundlich linear equation, log q = N log C + log KF, for three sets of hypothetical data from Figure 44, with fixed KF and varying N. The regression equations for these lines are summarized in Table 8. ........................................................................................................ 88
Figure 46. Freundlich adsorption isotherms, q=KFCN, for three sets of hypothetical data, with fixed N but with varying KF, as summarized in Table 8. ................ 90
Figure 47. Linear regression lines of the Freundlich linear equation, log q = N log C + log KF, for three sets of hypothetical data from Figure 46, with fixed N and varying KF, as summarized in Table 8. ................................................. 91
Figure 48. Linear regression lines of the linearly transformed Freundlich adsorption isotherms for adsorption of AsV by various Al:Fe hydroxides at pH 5, as summarized in Table 9. .................................................................................... 93
Figure 49. Linear regression lines of the linearly transformed Freundlich adsorption isotherms for adsorption of AsV by various Al:Fe hydroxides at pH 8, as summarized in Table 9. .................................................................................... 94
Figure 50. Linear regression lines of the linearly transformed Freundlich adsorption isotherms for adsorption of AsIII by various Al:Fe hydroxides at pH 5, as summarized in Table 10. .................................................................................. 95
Figure 51. Linear regression lines of the linearly transformed Freundlich adsorption isotherms for adsorption of AsIII by various Al:Fe hydroxides at pH 8 as summarized in Table 10. .................................................................................. 96
Figure 52. Adsorption envelopes of AsV in 0.1 M NaCl at an As:(Al+Fe) molar ratio of 0.05:1, at various Al:Fe molar ratios. .......................................................... 99
Figure 53. The hydrolysis species of Al3+ ion in equilibrium with gibbsite as a function of pH, calculated using thermodynamic constants tabulated in Lindsay (1979). .............................................................................................. 101
xiii
Page
Figure 54. Adsorption envelopes of AsIII in 0.1 M NaCl at an As:(Al+Fe) molar ratio of 0.05:1, at various Al:Fe molar ratios. ........................................................ 102
Figure 55. Adsorption isotherms of AsV in 0.1 molCa L-1 and 0.1 molNa L-1 at pH 5 as
affected by counterion and Al:Fe molar ratio. ............................................... 105
Figure 56. Adsorption isotherms of AsV in 0.1 molCa L-1 and 0.1 molNa L-1 at pH 8 as
affected by counterion and Al:Fe molar ratio. ............................................... 106
Figure 57. Adsorption envelopes of AsV in 0.1 molCa L-1 and 0.1 molNa L-1 as affected
by counterion and Al substitution at 0.1:1 As:(Al+Fe) molar ratio. .............. 108
Figure 58. Adsorption envelopes of AsV in 0.01 molCa L-1 and 0.01 molNa L-1 as
affected by counterion and Al substitution at 0.1:1 As:(Al+Fe) molar ratio. 109
Figure 59. Adsorption envelopes of AsV in 0.001 molCa L-1 and 0.001 molNa L-1as
affected by counterion and Al substitution at 0.1:1 As:(Al+Fe) molar ratio. 110
Figure 60. Adsorption envelopes of AsV in 0.1 molCa L-1 and 0.1 molNa L-1 as affected
by counterion and Al substitution at 0.025:1 As:(Al+Fe) molar ratio. .......... 111
Figure 61. Adsorption envelopes of AsV in 0.01 molCa L-1 and 0.01 molNa L-1 as
affected by counterion and Al substitution at 0.025:1 As:(Al+Fe) molar ratio. ............................................................................................................... 112
Figure 62. Adsorption envelopes of AsV with the 0:1 Al:Fe hydroxide at 0.1:1 As:(Al+Fe) molar ratio as affected by Ca counterion concentration. ............ 114
Figure 63. Adsorption envelopes of AsV with the 2:8 Al:Fe hydroxide at 0.1:1 As:(Al+Fe) molar ratio as affected by Ca counterion concentration. ............ 115
Figure 64. Adsorption envelopes of AsV with the 0:1 Al:Fe hydroxide at 0.025:1 As:(Fe+Al) molar ratio as affected by Ca counterion concentration. ............ 116
Figure 65. Adsorption envelopes of AsV with the 2:8 Al:Fe hydroxide at 0.025:1 As:(Fe+Al) molar ratio as affected by Ca counterion concentration. ............ 117
Figure 66. Adsorption envelopes of AsV with the 0:1 Al:Fe hydroxide at 0.1:1 As:(Fe+Al) molar ratio as affected by Na counterion concentration............. 118
Figure 67. Adsorption envelopes of AsV with the 2:8 Al:Fe hydroxide at 0.1:1 As:(Fe+Al) molar ratio as affected by Na counterion concentration............. 119
Figure 68. Adsorption envelopes of AsV with the 0:1 Al:Fe hydroxide at 0.025:1 As:(Fe+Al) molar ratio as affected by Na counterion concentration............. 120
xiv
Page
Figure 69. Adsorption envelopes of AsV with the 2:8 Al:Fe hydroxide at 0.025:1 As:(Fe+Al) molar ratio as affected by Na counterion concentration............. 121
Figure 70. Cation bridging by Ca2+................................................................................. 123
Figure 71. The diffuse double layers of Ca2+ vs. Na+. .................................................... 123
Figure 72. Desorption envelopes of AsV with sodium phosphate solution at 375 : 0.05 : 1 P:As:(Al+Fe) molar ratio, as affected by Al:Fe molar ratios. . 127
Figure 73. Desorption envelopes of AsIII with sodium phosphate solution at 375 : 0.05 : 1 P:As:(Al+Fe) molar ratio, as affected by Al:Fe molar ratios. . 128
Figure 74. Desorption envelopes of AsV and AsIII with sodium phosphate solution at 375 : 0.05 : 1 P:As:Fe molar ratio with the 0:1 Al:Fe hydroxide................... 130
Figure 75. Desorption envelopes of AsV and AsIII with sodium phosphate solution at 375 : 0.05 : 1 P:As:(Fe+Al) molar ratio with the 2:8 Al:Fe hydroxide.......... 131
Figure 76 . Desorption envelopes of AsV and AsIII with sodium phosphate solution at 375 : 0.05 : 1 P:As:(Fe+Al) molar ratio with the 5:5 Al:Fe hydroxide.......... 132
1
INTRODUCTION
Arsenic (As) introduced by natural processes or human activities can result in the
contamination of water. Arsenic is sufficiently toxic that its removal from contaminated
water is necessary if the water is to be consumed by humans. Many areas worldwide are
facing serious health problems due to As in drinking water, and the U.S. is not an
exception (Nriagu, 2002).
In October of 2001, the United States Environmental Protection Agency (USEPA)
lowered the maximum contaminant level (MCL) of As in municipal drinking water to 10
µgAs L-1 from the previous 50 µgAs L-1 standard (USEPA, 2001). Due to this new
regulation, the concentration of As in sludge and volume of As-containing sludge from
water-treatment plants are expected to increase as more As is removed from water. This
regulation has created demands for improved methods to remove As from water and to
control As in sludge.
Al or Fe hydroxides are used traditionally as adsorption agents in water-treatment
systems (Hammer and Hammer, 2001; Hering et al., 1996). Fe hydroxide is usually
considered to be a superior As adsorbent compared to Al hydroxide (Cheng et al., 1994;
Edwards, 1994; Gulledge and O'Connor, 1973; Hering et al., 1997; Tokunaga et al.,
1999); however, Fe and As compounds in contaminated residue from water treatment
can transform into soluble forms due to the redox processes involving Fe (Meng et al.,
2001), while solubilities of Al-hydroxide minerals are not as strongly affected by redox
processes. Al3+ substituted Fe3+ hydroxide might be able to combine the individual
This thesis follows the style and format of the Soil Science Society of America Journal.
2
advantages that the Al and Fe systems offer. An improved understanding of As
chemistry in mixed Al:Fe hydroxides could lead to improved methods of As treatment
and waste management. In addition, an improved understanding of As retention on Al3+
substituted Fe3+ hydroxides is necessary to fully understand As retention in soil, since the
majority of soil Fe hydroxides are known to be Al substituted (Schwertmann and Taylor,
1989).
3
OBJECTIVES
The purpose of this study is to examine the potential application of mixed Al:Fe
hydroxides in water treatment and residual stabilization by studying:
i. the mineralogy of Al3+ substituted Fe3+ hydroxides and their stabilities against
transformation, and
ii. the comparison of AsV and AsIII adsorption/desorption behavior on mixed Al:Fe
hydroxides as affected by Al substitution level, pH , and counterion (Ca2+ versus
Na+).
4
LITERATURE REVIEW
Arsenic in the Environment
The natural occurrence of As is typically associated with igneous and sedimentary
rocks either containing or derived from sulfidic compounds, geothermal areas, and fossil
fuels (Cullen and Reimer, 1989). Anthropogenic sources of As include by-products of
mining, metal refining, fossil fuels, and agriculture (Cullen and Reimer, 1989).
Extensive agricultural use of As in the U.S. has contributed to widespread contamination
of the environment (Nriagu, 2000). The agricultural utilization of As has decreased
drastically; however, traces of As can be found in food, water, air, and soil (Nriagu,
2000).
Chemical Properties of Arsenic
Arsenic has several possible oxidation states (-3, 0, +3, and +5), and its speciation
is strongly influenced by redox potential. Both inorganic and organically-bound As are
found in natural ecosystems; however, inorganic As species dominate in most aqueous
systems (Francesconi and Kuehnelt, 2002). Dissolved inorganic As exists mostly as AsIII
(arsenite) or AsV (arsenate) oxyanions in natural systems. AsV usually dominates under
oxidizing conditions, and AsIII is stable under reducing conditions (Cherry et al., 1979;
Smedley and Kinniburgh, 2002). Nonetheless, AsIII and AsV often coexist in both
reduced and oxidized environments due to the relatively slow kinetics of transformation
between oxidation states. Transformation of AsIII to AsV or vice versa can be either
5
abiotically or biotically mediated because the half cell potential of the AsV/AsIII couple is
within the redox-potential range of natural environments (Inskeep et al., 2002). For
example, the logKo of the AsV/AsIII redox couple resides between those of NO3-/N2 and
Fe(OH)3/Fe2+. The logKo values indicate that AsV will be reduced after NV, but before
FeIII (Table 1).
In addition to redox potential, pH influences the predominant inorganic As
species in aqueous systems (Table 2, Figure 1, and Figure 2). The pKa values indicate
that inorganic AsIII exists predominately as H3AsO3o and inorganic AsV exists as H2AsO4
-
and HAsO4 2- in most natural aqueous environments (Sadiq, 1997). Both AsIII and AsV
must be considered in the design of effective wastewater-treatment and waste-
management systems. The speciation of As in soil is spatially variable and seasonally
dependent, because pH, organic matter, biological activity, and redox potential, which are
also spatially and seasonally variable, influence the localized distribution of As species
(Inskeep et al., 2002; Masscheleyn et al., 1991b).
6
Table 1. Selected reduction half-reactions and thermodynamic constants (Sparks, 2003).
Half-reaction logKo
1/5NO3- + e- + 6/5H+ = 1/10N2 + 3/5H2O 21.1
1/2AsO43- + e- + 2H+ = 1/2AsO2
- + H2O 16.5
Fe(OH)3 + e- + 3H+ = Fe2+ + 3H2O 15.8
Table 2. The pKa values of AsIII (H3AsO3) and AsV (H3AsO4) (Wagman et al., 1982).
pKa1 pKa2 pKa3
AsIII 9.22 12.13 13.4
AsV 2.2 6.97 11.53
7
0 2 4 6
Spe
cies
, %
0
20
40
60
80
100
AsO33-H2AsO3
-
Figure 1. Speciation of AsIII as a function of p
H3AsO3º
pH
8 10 12 14
HAsO32-
H.
8
pH
0 2 4 6 8 10 12 14
Spe
cies
, %
0
20
40
60
80
100
AsO43-HAsO4
2-H3AsO3º H2AsO4-
Figure 2. Speciation of AsV as a function of pH.
9
Arsenic Toxicity
Epidemiological evidence indicates that As intake is detrimental to humans
(Smith et al., 2002a). Human response to As intake is highly variable as a result of
numerous intrinsic or extrinsic factors such as As dose, genetic variants, nutritional status,
age, pre-existing health conditions, and recreational habits (Anawar et al., 2002; NRC,
2001). Both cancer and noncancer health effects have been observed due to ingestion of
As (NRC, 2001). The proposed mode of action for As carcinogenicity is through
induction of chromosomal aberration without direct interaction with DNA (NRC, 2001).
Chronic exposure of As can cause skin, liver, kidney, bladder, and lung cancers (NRC,
2001; Smith et al., 1992). Inhibition of cellular respiration is known to be the
predominant mode of action for the noncancer effects of As (NRC, 2001). Several
symptoms of acute non-cancerous As-related illnesses include fever, anorexia,
hepatomegaly, melanosis, and cardiac arrhythmia. Neurotoxicity of both the peripheral
and central nervous system is observed as a result of chronic exposure of As (Goyer and
Clarkson, 2001).
The main source of inorganic As ingestion by humans is drinking water (Smith et
al., 1992); therefore, the assurance of safe drinking water is critical. Effective water
treatment is essential, because, at the previous MCL (50 µgAs L-1), human health could be
affected. Risk assessment by Smith et al. (1992) showed that the cancer mortality risk
was as high as 13 per 1000 persons from the lifetime ingestion of 1 L/day of water
containing 50 µgAs L-1.
10
Critical Problems
In order to achieve the new MCL, some water-treatment plants will need to
upgrade or install new treatment systems for effective As removal. The increase in cost
to meet the new As standard in the U.S. is expected to be approximately $200 million
annually (Smedley and Kinniburgh, 2002). Arsenic removal from water is challenging
because of the complex reaction of As.
Not only As removal, but also disposal and stabilization of the residual materials
generated from removing As, present technical challenges. Arsenic in sludge might be
remobilized due to possible change in pH and redox potential, especially when Fe is used
during the coagulation process. Meng et al. (2001) studied the effect of reductive
transformations of AsV and FeIII on As mobility from sludge generated by coprecipitation
with FeCl3. Upon reduction of the sludge, soluble AsIII and AsV concentrations were
increased, as reactive sites of the Fe hydroxide in the sludge were decreased as a result of
dissolution. The reactive surface sites of Fe hydroxide controlled the adsorption and
solubility of AsIII and AsV in the sludge. The results of this study indicate that the
biological reduction of sludge at land disposal sites might create a problem.
Bonding Mechanisms of Arsenic
The mode of As bonding on Fe hydroxides has been examined by extended X-ray
absorption fine-structure spectroscopy (EXAFS) as well as Fourier transform infrared
(FTIR) spectroscopy by various scientists (Fendorf et al., 1997; Goldberg and Johnson,
2001; Harrison and Berkheiser, 1982; Manceau, 1995; Manning et al., 1998; Sun and
Doner, 1996; Waychunas et al., 1993). Waychunas concluded from EXAFS studies that
AsV predominantly forms inner-sphere bidentate complexes on the ferrihydrite surface
11
(Figure 3), although about 30 % of all As-O-Fe complexes were monodentate complexes.
A bidentate-bridging bond of AsV on freshly prepared hydrous Fe oxide was also
observed by infrared spectroscopy (Harrison and Berkheiser, 1982). Goldberg and
Johnson (2001) utilized Raman and FTIR spectroscopy to detect inner-sphere
complexation of AsV and AsIII with amorphous Fe oxides; however, outer-sphere
complexation of AsIII was also detected. EXAFS results have shown that AsIII also forms
inner-sphere, bidentate binuclear-bridging complexes at the goethite surface (Manning et
al., 1998).
EXAFS results have indicated that both AsIII and AsV form inner-sphere bidentate
binuclear complexes with γ-Al2O3 (Arai et al., 2001). XANES spectra indicated the
formation of both inner-sphere and outer-sphere complexes of AsIII on γ-Al2O3. As pH
was increased and ionic strength was decreased, outer-sphere AsIII complexation
increased (Arai et al., 2001). Goldberg and Johnson (2001) observed in their Raman and
FTIR spectroscopy study that AsV forms inner-sphere complexes, and conversely that
AsIII forms only outer-sphere complexes with amorphous Al oxide.
O
OOH Fe
As
O Fe
Figure 3. Schematic illustration of the bidentate binuclear surface structure of AsV on Fe hydroxide.
12
Macroscopic Adsorption Behavior of Arsenic
Adsorption of anions on metal oxide has been of great interest because of the
importance of their removal during water treatment (Anderson et al., 1976). These
adsorption processes are pH dependent, and protons in solution and on the colloid surface
impact the rate of ligand exchange. Protons are available on the oxide surface below the
point of zero charge (pzc) or they could originate from free H+ at low pH or dissociation
of the conjugate acid of an inorganic or organic anion (Hingston et al., 1971).
An understanding of As adsorption on poorly crystalline Al and Fe hydroxides is
important because the solubility of As in natural environments is strongly influenced by
reactions at these highly reactive surfaces. In addition, poorly crystalline Al and Fe
hydroxides are expected to form during the coagulation processes of wastewater
treatment with Fe salts [e.g., FeCl3 and Fe2(SO4)3] and Al salts [e.g., Al2(SO4)3] (Hammer
and Hammer, 2001).
Fe hydroxide has a high affinity for As, but the reaction is highly dependent on
pH and the oxidation state of As. Studies have confirmed that AsV retention is usually
higher at low pH, and maximum AsV retention is usually achieved in the pH range of 4 to
5 (Dixit and Hering, 2003; Hingston et al., 1971; Hsia et al., 1992; Pierce and Moore,
1982; Raven et al., 1998). The Fe-hydroxide surface is positively charged below the pzc
(approximately at pH 8) due to increased protonation; therefore, conditions are favorable
for adsorption of negatively charged AsV species (Hsia et al., 1992). In the case of AsIII,
adsorption on a Fe hydroxide increases as pH increases up to the adsorption maximum at
approximately pH 8 to 10 (Ferguson and Anderson, 1974; Raven et al., 1998).
13
Negatively charged species dominate only above pH 9.2, due to the pKa of H3AsO3o, and
repulsive forces between AsIII and Fe-hydroxide surfaces are only appreciable at pH>9.
AsV adsorbs on Al hydroxides, whereas, AsIII is less readily adsorbed (Ferguson
and Anderson, 1974). The adsorption of AsV by Al hydroxide is dependent on pH
(Anderson et al., 1976; Ferguson and Anderson, 1974; Goldberg, 1986; Hingston et al.,
1971). Retention of AsV was greater at lower pH; however, AsV adsorption decreased at
pH<4.3 due to dissolution of amorphous Al hydroxide (Anderson et al., 1976).
In summary, the oxidation state of As and pH of the system are the most critical
factors affecting the inorganic As adsorption behavior on poorly crystalline Al and Fe
hydroxides.
Counterion Effect on Arsenic Adsorption
Counterions can strongly impact anion adsorption by soils and pure minerals
(Bowden et al., 1977). Increase in valency of the cation contributes to a less negatively
charged surface at pH values above the pzc of the mineral; therefore, under these
conditions, anion sorption increases with increasing valency of the counterion (Bowden
et al., 1977). An improved understanding of the counterion effect on As adsorption
would be valuable in the design of improved methods for water treatment.
The presence of Ca2+ compared to Na+ as the counterion has been reported to
significantly enhance the retention of AsV in soils, due to decrease in negative charge
character at the surfaces of soil minerals; however, AsIII adsorption by soil was little
affected by valency of the counterion (Smith et al., 2002b). Parks et al. (2003) observed
that the presence of dissolved Ca2+ enhanced the retention of As by Fe hydroxide and Al
hydroxide, and proposed that increased retention of As was due to reduced electrostatic
14
repulsion between the negatively charged surfaces and the As oxyanions. Reduction of
soluble As concentration was observed due to Ca arsenate precipitation at pH > 12 in a
mixture of CaCl2 and Na arsenate with no sludge; however, improved retention of As was
also observed at pH < 12 in a sludge suspension with lime. Although a considerable
amount of Ca2+ was adsorbed on the Fe hydroxide surface, the possibility of cation
bridging was eliminated as a reason for the differences, by diffuse layer modeling (Parks
et al., 2003). Jing et al. (2003) observed the reduced mobility of As from Fe sludge upon
cement treatment, due to the formation of Ca arsenate. The pH of the cement-treated, As-
containing sludge was reported to be 11.32. Ca-arsenate precipitation in the cement-
treated sample and inner-sphere complexation of AsV with Fe oxide in the non cement-
treated sample were detected by both FTIR and EXAFS (Jing et al., 2003). Formation of
Ca-arsenate solid was especially evident at pH > 7.3 in the mixture of Ca(OH)2 and the
AsV salt (Bothe and Brown, 1999). Depending on Ca/As molar ratio and pH, several
forms of Ca arsenate were formed (Table 3). The precipitation of Ca arsenate contributed
to reduce As leaching.
These results suggest that the use of Ca2+ for removal and stabilization of residual
materials could improve As retention.
Table 3. Ca arsenate hydrate precipitates (Bothe and Brown, 1999).
Solid-phase assemblage Ca/As pH
Ca4(OH)2(AsO4)2• 4H2O 2.2 ~ 2.5 12.23 ~ 12.54
Ca5(AsO4)3OH 1.9 ~ 1.67 12.63 ~ 9.77
Ca3(AsO4)2•32/3H2O
Ca3(AsO4)2• 41/4H2O 1.67 ~ 1.5 11.18 ~ 7.32
15
Effect of Ionic Strength on Arsenic Adsorption
A strong dependence of ionic strength is typically shown by anions forming outer-
sphere complexes (McBride, 1997). The adsorption of an anion (e.g., selenate) by outer-
sphere complexation is suppressed by competition with non-specifically adsorbed anions,
such as Cl- and NO3- (McBride, 1997). Inner-sphere complexes are less affected by ionic
strength Adsorption of anion (e.g., selenite) might be independent of ionic strength due to
strong bonding. Increase in anion adsorption (e.g., borate) with increase in ionic strength
has also been observed (McBride, 1997). This phenomenon is explained by the
contraction of the diffuse double layer, which allows the anion to more readily approach
the negatively charged oxide surface.
Goldberg and Johnson (2001) studied As adsorption on amorphous Fe and Al
oxides in 0.01 to 1.0 M NaCl by mean of adsorption envelopes. AsV adsorption by
amorphous Fe and Al oxides was independent of ionic strength, which is indicative of
inner-sphere complexation. AsIII adsorption by amorphous Al oxide decreased as ionic
strength was increased, but was only slightly dependent on ionic strength above pH 6. In
summary, AsIII adsorption on Al and Fe oxide surfaces was more strongly influenced by
ionic strength than is AsV adsorption.
Gupta and Chen (1978) studied the effect of ionic strength on adsorption of As by
activated alumina. Adsorption isotherms of AsIII at pH 6.5 to 8.5 and AsV at pH 6 to 7
were obtained in fresh water, diluted seawater, 0.67 M NaCl, and seawater. AsV and AsIII
adsorption on activated alumina decreased as the ionic strength was increased. For
example, AsV adsorption capacity of alumina was 4.11 mgAsV g-1 adsorbent in fresh water,
16
whereas, adsorption capacity was 0.81 mgAsV g-1 adsorbent in seawater. In addition, the
kinetics of AsIII and AsV adsorption by alumina was slower with higher ionic strength.
Goldberg and Johnson (2001) and Gupta and Chen (1978) obtained different
results with AsV adsorption on Al oxides as affected by ionic strength. Their results
indicate the complexity of the systems. The observed differences in adsorption might be
due to differences in experimental conditions such as adsorbent mineralogy, counterion,
pH, and As to adsorbent ratio. AsV adsorption by Al oxides should not be affected by
ionic strength, since the predominant mode of bonding between AsV and Al oxide is
known to be inner-sphere complexation according to the spectroscopic studies discussed
previously.
Adsorption Modeling
Adsorption reactions have been described using a variety of models such as the
Langmuir and Freundlich equations. The Langmuir equation was first developed in 1918
by Irving Langmuir to describe the adsorption of gaseous molecules on a homogeneous
planar surface, using several assumptions (Sparks, 2003). Most of these assumptions are
not met in heterogonous soil systems (Veith and Sposito, 1977). Although the Langmuir
equation has been widely used to model adsorption in soil systems, it should only be used
for qualitative purposes. In the Langmuir expression, the free energy of adsorption is
assumed to be independent of surface coverage (Reed and Matsumoto, 1993). Monolayer
coverage of the adsorbate at high C values and linear adsorption at low C values can be
described by the Langmuir isotherm. The Langmuir equation is presented as Equation
[1], and it can be transformed into a linear expression Equation [2] with 1/b as the slope
and 1/KLb as the intercept (Sparks, 2003). The KL and b parameters are usually
17
Q = (KLCb) / (1 + KLC) [1]
C/q = (1 / KLb) + (C / b) [2]
,where C = concentration of As in solution q = amount of As adsorbed b = calculated adsorption maximum KL = constant related to binding strength
considered to be a function of pH, ionic composition, and ionic strength. The KL and b
values are influenced by the electric double layer and the amphoteric behavior of the
surface (Reed and Matsumoto, 1993).
The Freundlich equation was first developed to describe gas-phase adsorption and
solute adsorption (Sparks, 2003). Unlike the Langmuir equation, the Freundlich equation
is not theoretically based. The Freundlich equation is presented as Equation [3], and it
can be transformed into a linear Equation [4] with N as the slope and log KF as the
intercept (Essington, 2004). In a broad sense, both KF and N are considered as constants
q = KFCN [3]
log q = N log C + log KF [4]
where, C = concentration of As in solution q = amount of As adsorbed KF = adjustable parameter N = adjustable parameter (0<N<1) characterizing the adsorption capacity (Yang, 1998). The constants KF and N are also
related to the strength of the adsorptive bond and bond distribution, respectively (Reed
and Matsumoto, 1993). It has been shown mathematically that N can be regarded a
measure of heterogeneity of adsorption sites (Yang, 1998). For example, surface site
18
heterogeneity increases as N approaches 0 (Essington, 2004). When N > 1, bond
energies increase with surface density (Reed and Matsumoto, 1993). When N < 1, bond
energies decrease with surface density. When N = 1, all surface sites are equivalent, and
the function is mathematically equivalent to the Langmuir isotherm with b approaching
infinity or KL << 1. Adsorption behavior with N < 1 is most common due to decreased
adsorption with increasing surface density (Reed and Matsumoto, 1993). In natural
systems, N is considered to be in between 0 and 1. The disadvantage of the Freundlich
equation is that it cannot be used to predict an adsorption maximum (Sparks, 2003). KF
and N values are also influenced by the electric double layer and the amphoteric behavior
of the surface.
Mineralogy of Mixed Al:Fe Hydroxides
The mineralogy of the adsorbent is a critical factor for As removal and
stabilization of residuals. Hematite and goethite are commonly found as products of
recrystallization of ferrihydrite (Schwertmann and Murad, 1983). Hematite formation
results from solid-phase transformation, while goethite formation occurs via dissolution
of ferrihydrite followed by reprecipitation, usually from Fe(OH)2+ and Fe(OH)4
-.
Formation of hematite as opposed to goethite is preferred at pH 7 to 8, where the
solubility of ferrihydrite is at the approximate minimum, whereas, maximum formation
of goethite as opposed to hematite has been reported at pH 4 and pH 12 (Schwertmann
and Murad, 1983). Preferential formation of hematite over goethite from Al-substituted
ferrihydrite has been reported by Schwermann et al. (2000), which indicates relatively
slow dissolution of the Al-substituted ferrihydrite.
19
Colombo and Violante (1996) synthesized a series of mixed Al:Fe hydroxides at
various Al:Fe molar ratios, by titrating mixtures of dissolved Fe(NO)3 and Al(NO)3 with
NaOH to pH 5, and studied the recrystallization of the products. Upon incubation of the
mixed Al:Fe hydroxides, changes in mineralogy were observed depending on the initial
Al:Fe molar ratio and temperature. Gibbsite, hematite, and goethite were detected as
products of the incubation; however, high stabilities against transformation were
observed at Al:Fe molar ratios of 2:8 to 5:5. This trend indicates a significant advantage
of Al:Fe hydroxides as an adsorbent in water treatment, since higher surface area is
favorable for anion adsorption, and the increased stability of the poorly crystalline phases
would decrease the potential for As release with time.
Arsenic Removal by Coagulation of Al and Fe Salts
Several techniques, such as ion exchange, adsorption by activated alumina and
activated carbon, ultrafiltration, reverse osmosis, and precipitation with or adsorption by
metal oxides followed by coagulation, have been used for removal of As from waste
water (Leist et al., 2000). Coagulation by Al or Fe salts is commonly used to remove As
in conventional water-treatment plants (Cheng et al., 1994; Hammer and Hammer, 2001;
Hering et al., 1996).
Arsenic is removed from wastewater much more efficiently as AsV than as AsIII
(Cheng et al., 1994; Gupta and Chen, 1978; Hering et al., 1997; Tokunaga et al., 1999).
AsIII is not as effectively removed by Al compared to Fe systems; however, AsV can be
removed by coagulation with Al hydroxide (Tokunaga et al., 1999). Fe coagulation
compared to Al coagulation is generally more effective in removing As (Cheng et al.,
1994; Edwards, 1994; Gulledge and O'Connor, 1973; Hering et al., 1997; Tokunaga et
20
al., 1999). Batch studies of the removal of AsV by coagulation with Fe2(SO4)3 and alum
have suggested that AsV is more effectively removed by Fe2(SO4)3 than by alum;
however, a larger coagulant dose improved the removal of AsV in both scenarios (Cheng
et al., 1994; Gulledge and O'Connor, 1973). AsV removal by Fe coagulation was most
effective at pH < 7, and its removal was independent of pH between 5.5 and 7.0 (Cheng
et al., 1994; Gulledge and O'Connor, 1973). Removal of AsV by Al hydroxide was
highly pH dependent at pH < 7 (Cheng et al., 1994; Gulledge and O'Connor, 1973).
These observations are consistent with the adsorption studies of AsIII and AsV on Fe and
Al hydroxides discussed above.
Desorption of Arsenic by Phosphate
Phosphate and AsV have similar chemical properties, and compete for the binding
sites of Al and Fe hydroxides; therefore, a reduction of AsV retention on Al and Fe
hydroxides has been reported in the presence of phosphate (Hingston et al., 1971;
Jackson and Miller, 2000; Jain and Loeppert, 2000; Liu et al., 2001; Manning and
Goldberg, 1996; Violante et al., 2002). Because of this phenomenon, extraction by high
phosphate solution has been utilized to assess As in soil (Alam et al., 2001; Davenport
and Peryea, 1991; Woolson et al., 1973). In waste disposal sites, the presence of
phosphate can significantly impact leaching of As from residual materials due to this
phenomenon.
Desorption of As by phosphate is dependent on oxidation state of As, pH, and
adsorbent. The kinetics of AsIII and AsV desorption from goethite exhibited different
trends. AsIII desorption reached an approximate maximum within 4 h; however, AsV was
continuously desorbed up to 100 h (Loeppert et al., 2002). Liu et al. (2001) studied the
21
desorption of AsV by phosphate and desorption of phosphate by AsV from goethite at an
AsV to phosphate molar ratio of 1:1 in the pH range of 3.0 to 8.5. The efficiency of
phosphate desorption by AsV was higher than that of AsV desorption by phosphate at any
given pH, and the effect of pH on desorption was greater with phosphate desorption by
AsV. This result is indicative of a stronger affinity of AsV on goethite relative to that of
phosphate.
The efficiency of AsV desorption is also affected by adsorbent, since more
phosphate than AsV was adsorbed on goethite, whereas, more AsV than phosphate was
adsorbed on gibbsite at an AsV to P molar ratio of 1:1 at both pH 4 and 7 (Violante et al.,
2002). This result indicates that there could be a difference in As-release potential from
residual material generated from Al and Fe coagulation.
Arsenic Analysis by Flow-Injection Hydride-Generation Flame-Atomic-Absorption
Spectroscopy
Flow-injection hydride-generation atomic-absorption spectroscopy is a widely
accepted analytical technique to analyze As at trace levels. The method involves reaction
of the sample in an acid medium with sodium borohydride (NaBH4) to convert AsIII into
gaseous arsine (AsH3) as summarized in Equation [5] (Masscheleyn et al., 1991a).
Arsine is transported by an inert gas such as argon to an atomizer of an atomic-absorption
spectrophotometer, where gas-phase atoms are generated. AsV species are reduced to
AsIII when solution pH is less than 1, which is then converted to AsH3. Because AsV
must be reduced to AsIII before the formation of AsH3, the kinetics of AsH3 formation
from AsV is slower than that from AsIII. The formation of AsH3 from AsV versus AsIII is
pH dependent, since As species must to be fully protonated to allow the reduction
22
reaction for the formation of AsH3 (Carrero et al., 2001). The advantage of As analysis
by this technique is the low detection limit.
NaBH4 + H3AsO3 + HCl AsH3(g) + H3BO3 + H2 + NaCl [5]
23
MATERIALS AND METHODS
Chemicals
Reagent grade chemicals were used in all studies. AsV and AsIII were obtained
from Alpha Aesar (Ward Hill, MA) as As2O5 and As2O3, respectively. AsV stock
solution was prepared in gently heated deionized water. AsIII stock solution was prepared
under N2 atmosphere with a minimum amount of NaOH added to ensure complete
dissolution of As2O3 at room temperature.
Synthesis of Ferrihydrite and Al-substituted Analogs
A series of Al-substituted hydroxides were prepared at 0:1, 3:97, 1:9, 2:8, 3:7, 4:6,
5:5, and 1:0 Al:Fe molar ratios. The two-line ferrihydrite method of Schwertmann and
Cornell (1991) was used except ferric nitrate [Fe(NO3)3] and aluminum nitrate [Al(NO3)3]
were hydrolyzed using sodium hydroxide (NaOH) rather than potassium hydroxide
(KOH).
In order to examine the counterion effect, 0:1 and 2:8 Al:Fe hydroxides were
prepared using saturated calcium hydroxide [Ca(OH)2] or 0.1 M NaOH to adjust the
appropriate Al(NO3)3 : Fe(NO3)3 mixtures to pH 7 to 8, to obtain systems with calcium
(Ca) or sodium (Na) as the only counterion. The initial concentrations of the hydroxides
were 0.004 molAl+Fe L-1. The volumes of saturated Ca(OH)2 or 0.1 M NaOH solution
needed were recorded to determine the accurate concentrations of Ca and Na in the
systems. The saturated Ca(OH)2 solution was prepared under nitrogen (N2) atmosphere
at room temperature with boiled deionized water to avoid the formation of calcium
carbonate (CaCO3) in the presence of carbon dioxide (CO2). The Ca(OH)2 solution was
immediately filtered through a 0.2 µm nominal pore-size membrane filter to remove any
24
precipitated CaCO3. The concentration of Ca(OH)2 was determined by titration with
hydrochloric acid (HCl). The Ca(OH)2 solution was prepared as close to the time of
hydroxide synthesis as possible to minimize CaCO3 formation.
Characterization of Synthesized Ferrihydrite and Al-substituted Analogs
X-ray Diffraction
Each hydroxide in a Nalgene bottle was shell frozen with liquid nitrogen.
Samples were then freeze-dried. X-ray diffraction analyses were performed on front
loaded power mounts with graphite monochromatized CuKα radiation from a Philip’s X-
ray diffraction unit, using a 0.05º step collected for 5 s from 2 to 65 º2θ.
Transmission Electron Microscopy
The morphology and aggregation of the hydroxides were examined using
transmission electron microscopy, on a JEOL 2010 TEM. To prepare the samples for
examination, dilute suspensions of hydroxides were sonicated in an ice bath for 1 h, and
mounted on silicon grids, which were first treated with chloroform.
Point of Zero Salt Effect
The point of zero salt effect (PZSE) and charge characteristics of each product
were determined using a batch titration procedure (Van Raij and Peech, 1972). During
the PZSE determination, the hydroxides were suspended in 1, 0.01, and 0.0001 M NaCl.
The concentration of Al+Fe in the suspensions was fixed at 0.01 molAl+Fe L-1. The pH
values of separate samples were adjusted from 3 to 11 in 0.4 pH unit intervals, using HCl
and NaOH. Following 2 h equilibration on a platform shaker, samples were centrifuged,
and equilibrium pH values were obtained while purging with N2 gas. The PZSE curve
was formed using the amount of HCl or NaOH added on the y-axis and pH on the x-axis.
25
Following filtration, the supernatant was analyzed by AAS to determine the concentration
of dissolved Al and Fe to allow for correction of H+ and OH- consumption due to
dissolved Al and Fe species.
Stability of Ferrihydrite and Al-substituted Analogs
The tendency of the 0:1 and 2:8 Al:Fe hydroxides to transform into crystalline
materials, and the mineralogy of the precipitated phases were examined following
incubation at pH 4 and 10, using the procedure summarized in Figure 4. Samples were
prepared with no added AsV and with an As:(Al+Fe) molar ratio of 0.05:1, and the pH
adjusted by adding HCl or NaOH. Samples were then incubated at 70oC, and subsamples
were taken after 12, 24, 48, and 96 h. The solution of AsV was added to the subsamples
which were incubated without AsV, and deionized water was added to the samples which
were incubated with AsV to ensure equal concentrations of As, Al and Fe in the two sets
of samples. The subsamples were adjusted to pH 7.0 with HCl or NaOH, and they were
aged at room temperature for 2 h to allow adsorption of AsV. Samples were centrifuged,
filtered, and analyzed for total dissolved As by FI-HG-AAS. The residual hydroxides
were washed with deionized water, freeze dried and then analyzed using XRD. Samples
were also extracted in the dark for 2 h with 0.2 M ammonium oxalate at pH 3.0 to
determine the proportion of poorly crystalline Fe hydroxide (Loeppert and Inskeep,
1996).
26
s
Incubate with AsV at pH 4 and 10 at 70oC
Subsamples- 12, 24, 48, and 96 h
XRD analysis of residual, and ammonium oxalate extraction of residual at pH 3 in the dark
Figure 4. Experimental procedure for hydroxides.
Hydroxide
Incubate without AsV at pH 4 and 10 at 70oC
Subsamples- 12, 24, 48, and 96 h
Add AsV, then adjust pH to 7 Age at RT for 2 h.
Add DI water, then adjust pH to 7Age at RT for 2 h.
Centrifuge, filter, then total As analysis
determination o
Centrifuge, filter, then total As analysis
XRD analysis of residual, and ammonium oxalate extraction of residual at pH 3 in the dark
f stability of the 0:1 and 2:8 Al:Fe
27
Analysis
Arsenic was analyzed by FI-HG-FAAS using a Perkin Elmer AA400 atomic
absorption spectrophotometer (Perkin Elmer Corporation, Norwalk, CT), with an
electrodeless discharge lamp (EDL) as the source of radiation. NaBH4 (1.5 %) in 0.5 %
NaOH was used as the reductant, and 5 M HCl was used as the eluent during flow
injection, to convert the As species in solution to AsH3 (Samanta et al., 1999). Gaseous
AsH3 was separated from the aqueous eluent using an ice water cooling system and a
gas/liquid separator, and transported to a quartz cell in FAAS. An air-acetylene flame
was used with a 10-cm burner head. The atomized As was analyzed at 193.7 nm
wavelength. The As detection limit was 0.5 µgAs L-1 with a 95 % confidence level.
Analysis of Al and Fe was also conducted using the Perkin Elmer AA400 atomic
absorption spectrophotometer. Nitrous oxide-acetylene and air-acetylene flames were
used for Al and Fe analyses, respectively. The matrix of the standard solutions was
matched to that of the samples in all cases.
Arsenic Adsorption Isotherms
Adsorption isotherms are used to present the adsorbate/adsorbent relationship and
are often represented as plots of the quantity of adsorbate retained by solid adsorbent as a
function of the equilibrium concentration of that adsorbate at fixed pH and ionic strength
(McBride, 1994). The capacities of 0:1, 2:8, 5:5, and 1:0 Al:Fe hydroxides to adsorb
AsIII and AsV were studied by means of adsorption isotherms at pH 5 and 8. The
reactions were conducted in 0.1 M NaCl ionic strength buffer, as a batch experiment with
As:(Al+Fe) molar ratios ranging from 0.0125:1 to 0.5:1. The Al+Fe concentration was
fixed at 267 µmolAl+Fe L-1 and As concentrations ranged from 3 to 133 µmolAs L-1, in
28
order to achieve the desired range of As:(Al+Fe) molar ratios. The pH values of separate
samples were adjusted by adding HCl or NaOH, and each sample was brought to 30 mL
final volume. Following equilibration for 24 h on a rotary shaker, the samples were
centrifuged and filtered through 0.2 µm nominal pore-size membrane filters. Supernates
were analyzed for total As by FI-HG-FAAS.
Arsenic Adsorption Modeling
Langmuir Adsorption Isotherms
Adsorption isotherm data were evaluated using the Langmuir equation (Equation
[2]). C/q versus C was plotted, and linear regression analyses were performed. The
calculated As adsorption maximum, b, and the constant related to binding strength, KL,
were examined. Because there is a large potential of analytical error with higher
As:(Al+Fe) molar ratios, all of the points over 0.2 As:(Al+Fe) molar ratio were excluded
for the Langmuir calculations.
Freundlich Adsorption Isotherms
Adsorption isotherm data were also evaluated using the Freundlich equation
(Equation [4]). Log q versus log C was plotted, and linear regression analyses were
performed. The calculated empirical constants, N and KF, were examined. All of the
points over 0.2 As:(Al+Fe) molar ratio were excluded for the Freundlich calculations.
Arsenic Adsorption Envelopes
Adsorption envelope is used to evaluate the influence of pH on adsorption at ionic
strength and constant adsorbent and adsorbate concentrations. The effect of pH and Al
substitution on AsIII and AsV adsorption was examined using adsorption envelopes.
Adsorption envelopes of AsIII and AsV on 0:1, 2:8, 5:5, and 1:0 Al:Fe hydroxides were
29
obtained in 0.1 M NaCl ionic strength buffer, as a batch experiment at a As:(Al+Fe)
molar ratio of 0.05:1 (13.35 µmolAs L-1 and 267 µmolAl+Fe L-1). The pH values of
individual samples were adjusted between 3 and 11 at 0.4 pH unit intervals by adding
HCl or NaOH, and each sample was brought to 30 mL final volume with deionized water.
Following 24 h equilibration on a platform shaker, samples were centrifuged, and the pH
values of the supernate were obtained. The samples were filtered through 0.2 µm
nominal pore-size membrane filters and analyzed by FI-HG-FAAS. The adsorption
envelopes were plotted using the proportion of the total adsorbed As (%) on the y-axis
and pH on the x-axis.
Arsenic Adsorption as Affected by Counterion
Adsorption envelopes and adsorption isotherms of AsV on 0:1 and 2:8 Al:Fe
hydroxides in Ca and Na systems were obtained to examine the counterion effect. The
procedures discussed previously were used, except the hydroxides prepared with
Ca(OH)2 and NaOH were used to maintain exclusively Ca and Na systems. The
concentration of counterion, Ca2+ and Na+, were adjusted using Ca(NO3)2 and NaNO3 salt
solutions.
For the adsorption envelopes, the suspensions were fixed at 0.1, 0.01, and 0.001
molCa L-1 and molNa L-1 to examine the effect of counterion concentration. The
adsorption envelopes were obtained at both 0.025:1 and 0.1:1 As:(Al+Fe) molar ratios by
varying AsV concentration (3.35 and 13.35 µmolAs L-1, respectively) with fixed Al+Fe
concentration (133.5 µmolAl+Fe L-1). For adsorption isotherms, the suspensions were
fixed at 0.1 molCa L-1 and molNa L-1. Nitrate salts were used since Fe(NO3)3 and
Al(NO3)3 were used to prepare the hydroxides. The Al+Fe concentration was fixed at
30
267 µmolAl+Fe L-1, and As concentrations ranged from 3 to 133 µmolAs L-1 in order to
achieve the desired As:(Al+Fe) molar ratios. Separate samples were adjusted by adding
HNO3 or NaOH to obtain pH values within the range of 3 to 11 in 0.4 pH unit intervals
for adsorption envelopes and at pH 5 and 8 for adsorption isotherms. The amount of
NaOH used to adjust pH in Ca systems was considered to be insignificant.
Arsenic Desorption Envelopes
The effect of pH and Al substitution on AsIII and AsV desorption by competitive
ligand change with phosphate was examined by mean of desorption envelopes. Arsenic
was first adsorbed on 0:1, 2:8, 5:5, and 1:0 Al:Fe hydroxides in a 0.1 M NaCl ionic
strength buffer at an As:(Fe+Al) molar ratio of 0.05:1 (26.7 µmolAs L-1 and 534 µmolAl+Fe
L-1) for 24 h. AsV was adsorbed at pH 5.2, and AsIII was adsorbed at pH 8.5. These pH
values were used since the adsorption maxima of AsV and AsIII were found at
approximately pH 5.2 and pH 8.5, respectively, for all hydroxides in the adsorption
envelope study. Sub-samples were taken from each suspension before the addition of
phosphate to determine the amount of As adsorbed after 24 h. Following As adsorption,
desorption envelopes were obtained as a batch experiment. Ten milliliters of 0.2 M
sodium phosphate solution, with pH preadjusted from 3 to 11 in 0.4 pH unit intervals,
were added to each bottle containing 10 ml of As-treated hydroxide suspension.
Deionized water was added to a separate As-treated hydroxide suspension as a control to
evaluate whether desorption of As was due to mechanical agitation. The total
concentration of As during the desorption reaction was 13.35 µmolAs L-1, and sodium
phosphate concentration was 0.1 M (1:7491 As:P molar ratio). Each sample was allowed
to react for 24 h on a rotary shaker. Upon completion of the reaction, samples were
31
centrifuged, and the pH values of the supernatant solutions were obtained. The samples
were filtered through 0.2 µm nominal pore-size membrane filters, and analyzed by FI-
HG-FAAS. The desorption envelopes were plotted using the percent of As desorbed on
the y-axis and pH on the x-axis.
32
RESULTS AND DISCUSSION
Characterization of Synthesized Ferrihydrite and Al-substituted Analogs
X-ray Diffraction
Differences in XRD patterns of synthesized Al:Fe hydroxides were observed with
the varying Al:Fe molar ratios. The 2:8 Al:Fe hydroxide resulted in an XRD pattern
almost identical to that of ferrihydrite, except the peaks were boarder (Figure 5). The
peak widths at half height were 7.55, 8.35, and 9.37 º2θ for the 0:1, 2:8, and 3:7 Al:Fe
hydroxides, respectively. XRD line broadening can be indicative of both smaller crystal
size and a reduction in long-range order of the materials. The XRD pattern of the 2:8
Al:Fe hydroxide suggests that this material has a smaller particle size than that of the 0:1
Al:Fe hydroxide. A smaller particle size of goethite (α-FeOOH) was observed with
increasing Al substitution, by both XRD and TEM analysis (Fey and Dixon, 1981;
Schulze and Schwertmann, 1984). The ionic radius of Al3+ (0.53 Å) is slightly smaller
compared to that of Fe3+ (0.65 Å); therefore, isomorphous substitution of Al3+ for Fe3+
would result in a decrease in average size of the unit cell (Schulze, 1984), which would
result in peak shifts in the XRD pattern. In the current study, there was a small tendency
toward shift to higher º2θ at the higher Al contents. The broad peak at approximately 62
º2θ also shifted towards higher º2θ with the higher Al contents. With the 4:6 and 5:5
Al:Fe hydroxides, there was an indication of the peak splitting of the 35 º2θ peak, which
is probably due to the presence of a separate Al-rich phase at the higher Al contents.
Heterogeneous distribution of Al3+ within the structure could also contribute to peak
broadening. Small peaks of gibbsite [γ-Al(OH)3], which increased in size with time,
33
0
50
100
150
200
250
0 5 10 15 20 25 30 35 40 45 50 55 60 65Degrees 2-theta
Cou
nts
per S
econ
d
0:1
2:8
3:7
4:6
5:5
B,G
B,GBG G
G G
B,G
G G G G G G
B --- bayeriteG --- gibbsite
Figure 5. XRD patterns of freshly prepared hydroxides at various Al:Fe molar ratios.
34
were observed with the 3:7 Al:Fe hydroxide (Figure 6). After 1 month aging at 2ºC, a
white precipitate was observed in the storage bottle. Only poorly crystalline material was
detected with the 0:1 and 2:8 Al:Fe hydroxides, even upon aging. Bayerite and gibbsite
were both identified as products of the synthetic systems with Al:(Al+Fe) molar ratios
greater than 0.4 (Figure 5 and Figure 7). At 0.4 Al:(Al+Fe) molar ratio, only a small
peak of bayerite was observed at 40.7 º2θ (Figure 5), which indicates that the
predominant crystalline product was gibbsite. Although the 1:0 Al:Fe hydroxide
contained significant amounts of both bayerite and gibbsite, the broad background peaks
were indicative of poorly crystalline Al hydroxide (Figure 7). Crystalline Fe hydroxide
minerals were not found in any of the synthesized materials. In summary, Al
incorporates quantitatively into the poorly crystalline ferrihydrite structure, with no
evidence of a crystalline Al hydroxide phase, up to approximately 0.20 Al:(Al+Fe) molar
ratio.
Gibbsite and bayerite are composed of identical structural units, that is, two
planes of close-packed OH- with Al3+ between them (Hsu, 1989). These Al(OH)3 sheets
are held together by hydrogen bonding. Two-thirds of the octahedral sites are filled with
Al3+, to form a planar hexagonal ring structure, in which each Al3+ shares six OH- with
three other Al3+ ions, and each OH- is bridged between two Al3+ ions. In gibbsite, one-
half of the OH- groups point away, perpendicularly, from the octahedral sheet, while half
of the OH- groups on adjacent sheets reside directly opposite from the perpendicular OH-
35
0
20
40
60
80
100
120
140
0 5 10 15 20 25 30 35 40 45 50 55 60 65
Degrees 2-theta
Cou
nts
per S
econ
d
Aged 3:7
Fresh 3:7
B,G G
G
B --- bayeriteG --- gibbsite
BB,G
Figure 6. XRD patterns of fresh 3:7 Al:Fe hydroxides and 3:7 Al:Fe hydroxides aged at 2ºC for 1 month.
36
0
100
200
300
400
500
600
700
0 5 10 15 20 25 30 35 40 45 50 55 60 65Degrees 2-theta
Cou
nts
per S
econ
d
5:5
1:0
B --- bayeriteG --- gibbsite
B,G
B,G
B,G B,G B GBG
GG
G G G G G G
B
Figure 7. XRD patterns of 1:0 and 5:5 Al:Fe hydroxides.
37
groups but parallel to the basal plane. The OH- planes of gibbsite have an AB-BA-AB-
BA stacking arrangement (Figure 8) (Wefers and Bell, 1972). Gibbsite has zero net
permanent charge because there is no significant isomorphous substitution of Al3+ by
divalent cations (Huang et al., 2002). In bayerite, the perpendicular OH- groups in one
plane lie in the depression of the adjacent plane (Wefers and Bell, 1972). As a result, the
crystal lattice of bayerite is composed of layers of OH- with an AB-AB-AB stacking
arrangement, as opposed to the AB-BA-AB-BA sequence for gibbsite (Figure 9).
The hue of the hydroxides became less red (10R to 5YR), and value and chroma
increased as Al substitution increased (Figure 10). In the case of hematite (α-Fe2O3),
both value and chroma increased as Al substitution increased; however, hue was
independent of Al substitution (Kosmas et al., 1986). For goethite, value and hue
decreased as Al substitution increased, although chroma was independent of Al
substitution (Kosmas et al., 1986). The variation in color of the Fe oxides and hydroxides
are often indicative of differences in mineral structure (Schwertmann and Taylor, 1989).
Darker colors (low values) are often exhibited with condensed masses (Schwertmann and
Taylor, 1989), which indicates that the 0:1 Al:Fe hydroxide was likely to be most
condensed compared to the Al substituted hydroxide.
38
H A B B A A B
O
Figure 8. Structure of gibbsite.
A B A B A B
H O
Figure 9. Structure of bayerite.
0:1 2:8 3:7 4:6 5:5
10R2.5/1 10R2.5/2 2.5YR2.5/4 2.5YR3/6 5YR4/6
Figure 10. Colors of freeze dried ferrihydrite and its Al-substituted analogs.
39
Transmission Electron Microscopy
In TEM micrographs, differences in aggregation were observed with change in
Al:Fe molar ratio. The 0:1 Al:Fe hydroxide was observed only as aggregates of varying
density (Figure 11 and Figure 12). The 2:8 Al:Fe hydroxide was generally more
dispersed than the 0:1 Al:Fe hydroxide (Figure 11 versus Figure 13). Low-density
regions were observed with the 2:8 Al:Fe hydroxide (Figure 13 and Figure 14); however,
high-density aggregates similar to those found with the 0:1 Al:Fe hydroxide were also
observed (Figure 13). Hexagonal gibbsite and pyramidal bayerite crystals along with
small aggregates of poorly crystalline hydroxide were observed with the 5:5 Al:Fe
hydroxide (Figure 15 and Figure 16-a). The aggregates with the 5:5 Al:Fe hydroxide
were generally smaller and less dense than those observed with the 0:1 and 2:8 Al:Fe
hydroxides. Crystalline products in the 5:5 Al:Fe hydroxide were identified as gibbsite
(hexagonal plates) and bayerite (triangular pyramidal) (Hsu, 1989). Smaller gibbsite and
bayerite crystals were observed with the 1:0 Al:Fe hydroxide than with the 5:5 Al:Fe
hydroxide (Figure 15-a versus Figure 16-b). Determination of the crystal size was
difficult because Al hydroxides in the 1:0 Al:Fe hydroxide sample were unstable under
the TEM electron beam, and the image was distorted due to the evolution of water vapor.
Aggregates of poorly crystalline Al hydroxide were also found in the 0:1 Al:Fe
hydroxide.
40
igure 11. TEM micrographs of the 0:1 Al:Fe hydroxides. (a) Aggregates of 0:1 Al:Fe hydroxide. (b) Dense aggregates of 0:1 Al:Fe hydroxide.
a
200 nm
b
200 nm
F
41
100 nm
a
b
100 nm
Figure 12. TEM mydroxide (magnified picture of Figure 11-a). (b) Dense aggregates of 0:1 Al:Fe ydroxide (magnified picture of Figure 11-b).
icrographs of the 0:1 Al: Fe hydroxides. (a) Aggregates of 0:1 Al:Fe hh
42
a
b
200 nm
200 nm
Figure 13. TEM m
hydroxide w
icrographs of the 2:8 Al:Fe hydroxides. (a) Dense aggregates of the 2:8 Al:Fe hydroxide and porous surrounding. (b) Dense aggregates of 2:8 Al:Fe
ith dispersed aggregates.
43
a
Figure 14. TEM m ates of 2:8 Al:Fe hydroxide (m(magnified from
icrographs of the 2:8 Al:Fe hydroxides. (a) Aggregagnified from Figure 13-a). (b) Aggregates of 2:8 Al:Fe hydroxide
Figure 13-b).
b
100 nm
100 nm
44
a Gibbsite
Aggregate
Bayerite
400 nm
60 nm b
Figure 15. TEM mall aggregates of the 5:5 Al:Fe hydroxide. (b) Magnified image of bayerite crystal
und in the 5:5 Al:Fe hydroxide.
icrographs of the 5:5 Al:Fe hydroxides. (a) Gibbsite, bayerite, and smfo
45
igure 16. TEM m age f gibbsite crystal found in the 5:5 Al:Fe hydroxide. (b) Gibbsite, bayerite, and aggregate f Al hydroxide found in the ydroxide.
100 nm
a
b
400 nm
Gibbsite
Bayerite Aggregate
Fo
icrographs of the 5:5 and 0:1 Al:Fe hydroxides. (a). Magnified im
o 0:1 Al:Fe h
46
The TEM study further confirmed that 0:1 and 2:8 Al:Fe hydroxides are poorly
e tendency of the poorly crystalline
hydroxides to aggregate, quantitative determination of surface area is challenging.
Multiple phases (gibbsite, bayerite, and poorly crystalline product) were observed in the
5:5 and 1:0 Al:Fe hydroxides, which confirms the XRD data.
Point of Zero Salt Effect
The PZSEs of 0:1, 2:8, 5:5, and 1:0 Al:Fe hydroxides were approximately 7.6, 8.2,
.7, and 8.9, respectively (Figures 17 – 20). Titration curves followed similar trends
regardless of Al:Fe molar ratio; however, pH values at PZSE increased as Al:Fe molar
ratio was increased. This trend indicates the reversal in the net charge of the surface
occurs at higher pH with higher Al:Fe molar ratio. The charge characteristics of the 1:0
Al:Fe hydroxide were not as strongly influenced by ionic strength compared to the other
hydroxides.
crystalline and highly aggregated; however, the TEM images revealed that the 2:8 Al:Fe
hydroxide was generally more dispersed. Due to th
8
47
y = -0.001x3 + 0.0225x2 - 0.197x + 0.6343R2 = 0.9926
y = -0.0028x3 + 0.0603x2 - 0.4283x + 0.9697R2 = 0.9688
y = -0.0017x3 + 0.0384x2 - 0.2959x + 0.7701R2 = 0.9969
-0.2
-0.15
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.253 4 5 6 7 8 9 10 11
pH
mol
H+
mol
Al+
Fe-1
0.0001 M0.01 M1 M
Figure 17. Titration curves for the 0:1 Al:Fe hydroxide at three ionic strengths.
48
y = -0.0036x3 + 0.0764x2 - 0.5735x + 1.5009R2 = 0.988
y = -0.0021x3 + 0.0472x2 - 0.375x + 1.0081R2 = 0.9972
y = -0.0038x3 + 0.0827x2 - 0.5967x + 1.4005R2 = 0.9875
-0.2
-0.15
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.253 4 5 6 7 8 9 10 11
pH
mol
H+
mol
Al+F
e-1
0.0001 M0.01 M1 M
Figure 18. Titration curves for the 2:8 Al:Fe hydroxide at three ionic strengths.
49
y = -0.0014x3 + 0.0327x2 - 0.2639x + 0.7024R2 = 0.996
y = -0.0034x3 + 0.0683x2 - 0.4733x + 1.1263R2 = 0.9831
y = -0.0071x3 + 0.1292x2 - 0.7905x + 1.5904R2 = 0.9944
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.33 4 5 6 7 8 9 10 11
pH
mol
H+
mol
Al+F
e-1
0.0001 M0.01 M1 M
Figure 19. Titration curves for the 5:5 Al:Fe hydroxide at three ionic strengths.
50
y = -0.0045x3 + 0.0975x2 - 0.7007x + 1.5573R2 = 0.9762
y = -0.004x3 + 0.0853x2 - 0.6075x + 1.3611R2 = 0.9462
y = -0.0046x3 + 0.0949x2 - 0.6592x + 1.4803R2 = 0.9837
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.33 4 5 6 7 8 9 10 11
pH
mol
H+
mol
Al+F
e-1
0.0001 M0.01 M1 M
Figure 20. Titration curves for the 1:0 Al:Fe hydroxide at three ionic strengths.
51
Stability of Ferrihydrite and Its Al-substituted Analogs
Effect of Al Substitution and pH
After 70oC incubation at pH 10 for 96 h without AsV, no poorly crystalline Fe
hydroxide remained in the 0:1 Al:Fe hydroxide, as detected by pH 3 ammonium-oxalate
extraction in the dark; however, 64 % of the poorly crystalline Fe hydroxide remained in
the 2:8 Al:Fe hydroxide (Table 4). Also, 53 % of the Fe hydroxide was poorly crystalline
in the 0:1 Al:Fe hydroxides following incubation at pH 4, although transformation of the
2:8 Al:Fe hydroxide at pH 4 into a crystalline hydroxide was not evident by pH 3
ammonium-oxalate extraction in the dark (Table 4). The proportion of crystalline Fe
hydroxide to poorly crystalline phase is known to increase with increasing incubation pH
(Schwertmann and Murad, 1983). The results of ammonium oxalate extraction indicate
that differences in stability of the originally synthesized poorly crystalline phases against
transformation into crystalline phases are influenced by Al substitution.
Table 4. Percent of ammonium oxalate extractable (AOE) Fe to total Fe following incubation at 70 oC for 96 h.
Incubated with As Incubated without As
Treatment
0:1 Al:Fe pH 10 97 00:1 Al:Fe pH 4 95 532:8 Al:Fe pH 10 99 642:8 Al:Fe pH 4 100 100
Ammonium oxalate extractable Fe/Total Fe%
52
The influence of Al substitution and incubation pH on incubation product
mineralogy was also determined by XRD (Figure 21). Following incubation at pH 10 for
96 h, hematite and goethite were detected by XRD analysis of the 0:1 Al:Fe hydroxide;
however, only hematite was detected with the 2:8 Al:Fe hydroxide. The broad
background peak of the incubated 2:8 Al:Fe hydroxide is an indication that the sample
still contained considerable poorly crystalline hydroxide. Following pH 4 incubation of
the 0:1 Al:Fe hydroxide, hematite and goethite were detected, although no crystalline
material was found in the incubation product of the 2:8 Al:Fe hydroxide. XRD analyses
of the final products were compatible with the result of the pH 3 ammonium-oxalate
extractions in the dark (Table 4 and Figure 21).
The formation of goethite and hematite from ferrihydrite at pH 10 versus the
formation of only hematite at pH 4 is influenced by Al substitution and pH, because the
mechanisms of formation of these phases differ. Goethite is formed by dissolution of
ferrihydrite, followed by reprecipitation, usually from Fe(OH)2+ and Fe(OH)4
-; however,
hematite is formed by internal rearrangement and dehydration within the ferrihydrite
(Schwertmann and Murad, 1983). The solubility of ferrihydrite is pH dependent, and the
53
0
100
200
300
400
500
600
0 10 20 30 40 50 60
Degrees 2-theta
Cou
nts
per S
econ
d
Hematite + Goethite
Hematite + Goethite
Hematite
A --- 0:1 pH 10B --- 0:1 pH 4C --- 2:8 pH 10D --- 2:8 pH 4
A
B
C
D
Figure 21. XRD patterns of the hydroxides incubated without AsV at 70 oC for 96 h.
54
minimum solubility of ferrihydrite occurs at pH 7.5 to 8.5 (Figure 22) (Lindsay, 1979).
Maximum formation of hematite as opposed to goethite has been reported at pH 7 to 8,
where the solubility of ferrihydrite is at an approximate minimum; maximum formation
of goethite as opposed to hematite has been reported at pH 4 and pH 12, where the
principal dissolved species are Fe(OH)2+ and Fe(OH)4
-, respectively (Schwertmann and
Murad, 1983). The formation of hematite only was observed with the 2:8 Al:Fe
hydroxide at pH 10, where goethite formation should be favored (Figure 21). This
phenomenon indicates that the 2:8 Al:Fe hydroxide might be less soluble compared to the
0:1 Al:Fe hydroxide (Schwertmann et al., 2000). Preferential formation of hematite over
goethite from Al-substituted ferrihydrite was also observed by Schwermann et al. (2000).
Kinetics of Mineral Transformation
The kinetics of mineral transformation was also affected by Al substitution and
pH. The hydroxides equilibrated at pH 4 transformed more slowly than those
equilibrated at pH 10 (Figure 23 versus Figure 24). The decrease in the rate of
ferrihydrite transformation with decrease in pH was also observed by Schwertmann and
Murad (1983). The relationship between the proportion of AsV not adsorbed and the
proportion of the crystalline Fe hydroxide is summarized in Figure 25. The proportion of
AsV not adsorbed increased as the surface area decreased, due to the formation of
crystalline hydroxides. In addition, the slower increase in AsV concentration in solution
with the 2:8 Al:Fe hydroxides relative to the 0:1 Al:Fe hydroxide indicates that the
mineral transformation was slower with the 2:8 Al:Fe hydroxide. Schwertmann et al.
(2000) also showed that the transformation of poorly crystalline Fe hydroxide was slower
when Al substitution of ferrihydrite was increased.
55
igure 22. The activities of hydrolysis species of Fe3+ in equilibrium with amorphous Fe
pH
3 4 5 6 7 8 9 10
log
activ
ity
-22
-20
-18
-16
-14
-12
-10
-8
-6
-4
Fe2(OH)24+
Fe3+
FeOH2+
Fe(OH)3
O
Fe(OH)4-
Total Fe
Fhydroxide as a function of pH, calculated using thermodynamic constants from Lindsay (1979).
56
Time, h
0 20 40 60 80 100
Equ
libriu
m A
s co
ncen
tratio
n in
sol
utio
n, %
0
20
40
60
80
100
0:1 Al:Fe2:8 Al:Fe
VFigure 23. The influence of incubation time on the proportion of added As not adsorbed. The incubation was at pH 10 and 70 oC. The AsV (0.05 mmolAs mmolAl+Fe
-1) was added to the suspension following incubation and was allowed to equilibrate at pH 7 and 23 ºC for 2 h.
57
Time, h
0 20 40 60 80 100
Equ
libriu
m A
s co
ncen
tratio
n, %
0
5
10
15
20
25
0:1 Al:Fe2:8 Al:Fe
FT
igure 24. The influence of incubation time on the proportion of added AsV not adsorbed. he incubation was at pH 4 and 70 oC. The AsV (0.05 mmolAs mmolAl+Fe
-1) was added to e suspension following incubation and was allowed to equilibrate at pH 7 and 23 ºC for
th2 h.
58
Crystalline Fe, %
0 20 40 60 80
As
rem
aini
ng in
sol
utio
n, %
0
20
40
60
80
100
pH 10, 70 oC incubationpH 4, 70 oC incubation
r2=0.998 Y=0.471X+0.220
r2=0.995 Y=1.03X-2.56
100
Figure 25. AsV remaining in solution from the final incubation products as a function of proportion of crystalline Fe not extracted by pH 3 ammonium oxalate from the products. The AsV was added to the suspension following incubation and was allowed to equilibrate at pH 7 and 23 ºC for 2 h.
59
Effect of Arsenic
The transformation of both 0:1 and 2:8 Al:Fe hydroxides into crystalline products
was retarded in the presence of AsV (Table 4). There are two possible factors by which
AsV might contribute to this overall relationship. Adsorbed AsV at the surface of nuclei
might poison the crystal growth of the transformation product of both 0:1 and 2:8 Al:Fe
hydroxides. Also, dissolution of ferrihydrite might be reduced by the presence of AsV
(Paige et al., 1996). In the latter case, transformation to goethite might be limited
because goethite formation is dependent on ferrihydrite dissolution. The presence of
coprecipitated or adsorbed AsV on amorphous Fe hydroxide slowed the rate of
transformation into crystalline products at pH 12 (Paige et al., 1996). This trend indicates
that the AsV-contaminated sludge would be more resistant to transformation than the non-
contaminated adsorbent.
The results of this study indicate that coprecipitation of Al during precipitation of
poorly crystalline Al:Fe hydroxide resulted in a product that was more resistant to
transformation into well crystalline goethite or hematite. In the water treatment scenario,
the use of Al with Fe during coagulation might contribute to the maintenance of higher
surface areas for As adsorption.
Arsenic Adsorption Isotherms
Effect of Mineralogy
There were differences in the adsorption of AsV on poorly crystalline hydroxides
(0:1 and 2:8 Al:Fe hydroxides) versus the 5:5 Al:Fe hydroxide at both pH 5 and 8 (Figure
26 and Figure 27). Furthermore, adsorption of AsV on the 1:0 Al:Fe hydroxide was
60
Equilibrium As, mmolAs L-1
0.00 0.02 0.04 0.06 0.08 0.10 0.12
As
adso
rbed
, mol
As m
olA
l+Fe
-1
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
0.18
0:1 Al:Fe2:8 Al:Fe5:5 Al:Fe1:0 Al:Fe
Figure 26. Adsorption isotherms of AsV at pH 5 on precipitated products of various Al:Fe molar ratio.
61
Equilibrium As, mmolAs L-1
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
As
adso
rbed
, mol
As m
olA
l+Fe
-1
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0:1 Al:Fe2:8 Al:Fe5:5 Al:Fe1:0 Al:Fe
Figure 27. Adsorption isotherms of AsV at pH 8 on precipitated products of various Al:Fe molar ratio.
62
substantially lower than that on the Fe-containing adsorbents, regardless of pH. This
difference in AsV adsorption is primarily due to the differences in surface area. The XRD
and TEM study revealed the presence of crystalline bayerite and gibbsite in the 5:5 and
1:0 Al:Fe hydroxides; therefore, a lower concentration of surface adsorption sites would
be expected with these hydroxides compared to the poorly crystalline 0:1 and 2:8 Al:Fe
hydroxides. The low reactivity of gibbsite and bayerite can be explained by their
relatively inert structures. Since there is no significant isomorphous substitution of Al3+
by divalent cation, the net permanent charge approaches zero in gibbsite and bayerite.
The only sites for As adsorption are the edge sites, because all OH- groups on the planar
surfaces are charge satisfied. Even at the edge site, one half of the OH- groups are charge
satisfied since they are doubly coordinated to two Al3+ ions. The only reactive sites are
the other half of the OH- groups, which are undercoordinated (Essington, 2004). In
addition, Al hydroxide has a strong tendency to grow in the X and Y plane, but crystal
growth is often limited in the Z direction, which also contributes to a lower concentration
of surface adsorption sites. The crystal growth of gibbsite and bayerite is influenced by
strong Al-OH-Al bonding within the layer structure and weak bonding between layers via
hydrogen bonding (Hsu, 1989). Because the poorly crystalline hydroxides have higher
concentrations of surface adsorption sites, it is predictable that higher concentrations of
As would be adsorbed. At both pH 5 and 8, AsV was quantitatively adsorbed by the 5:5
Al:Fe hydroxide up to approximately 0.025 As:(Fe+Al) molar ratio, although the 0:1 and
2:8 Al:Fe hydroxides adsorbed AsV quantitatively up to approximately 0.05 As:(Al+Fe)
molar ratio (Figure 28 and Figure 29). The 0:1 and 2:8 Al:Fe hydroxides exhibited a
63
Equilibrium As, mmolAs L-1
0.000 0.001 0.002 0.003 0.004 0.005 0.006
As
adso
rbed
, mol
As m
olA
l+Fe
-1
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0:1 Al:Fe2:8 Al:Fe5:5 Al:Fe1:0 Al:Fe
Figure 28. Adsorption isotherms of AsV at low equilibrium AsV concentrations at pH 5 on precipitated products of various Al:Fe molar ratio.
64
Equilibrium As, mmolAs L-1
0.000 0.002 0.004 0.006 0.008 0.010
As
adso
rbed
, mol
As m
olA
l+Fe
-1
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0:1 Al:Fe2:8 Al:Fe5:5 Al:Fe1:0 Al:Fe
Figure 29. Adsorption isotherms of AsV at low equilibrium AsV concentrations at pH 8 on precipitated products of various Al:Fe molar ratio.
65
higher capacity to adsorb AsV, regardless of pH, at any As:(Al+Fe) molar ratio (Figure 26
and Figure 27). A slightly higher retention of AsV was observed with the 0:1 Al:Fe
hydroxide than with the 2:8 Al:Fe hydroxide; however, the statistical significance was
uncertain.
AsV is usually considered to be more effectively removed by Fe than by Al
coagulation (Cheng et al., 1994; Edwards, 1994; Gulledge and O'Connor, 1973; Hering et
al., 1997; Tokunaga et al., 1999), even though the AsV adsorption mechanism is
predominately inner-sphere complexation with both Al and Fe oxyhydroxides (Arai et al.,
2001; Fendorf et al., 1997; Goldberg and Johnson, 2001; Harrison and Berkheiser, 1982;
Manceau, 1995; Manning et al., 1998; Sun and Doner, 1996; Waychunas et al., 1993).
The generally lower efficiencies of AsV removal in the case of Al hydroxide might have
been influenced by the relatively inert structure of the crystalline Al hydroxide minerals,
gibbsite and bayerite, which are likely produced during the coagulation processes.
AsIII was not quantitatively adsorbed regardless of adsorbent and pH (Figures 30 –
33). Adsorption of AsIII decreased as Al substitution was increased, and adsorption of
AsIII on the 1:0 Al:Fe hydroxide was negligible at all As:Al molar ratios, regardless of the
pH. This result indicates that AsIII is strongly adsorbed only by the Fe3+ ion within the Al
substituted Fe hydroxides. Because the sum of Fe plus Al present in all suspensions was
consistent throughout this study, less Fe was likely available at surface adsorption sites as
Al substitution was increased.
Weak affinity for AsIII adsorption on gibbsite and amorphous Al hydroxide and
relatively slow kinetics of AsIII adsorption by gibbsite have been reported previously
(Goldberg and Johnson, 2001; Weerasooriya et al., 2003). Affinity of AsIII by Al
66
Equlibrium As, mmolAs L-1
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
As
adso
rbed
, mol
As m
olA
l+Fe
-1
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0:1 Al:Fe2:8 Al:Fe5:5 Al:Fe1:0 Al:Fe
Figure 30. Adsorption isotherms of AsIII at pH 5 on precipitated products of various Al:Fe molar ratio.
67
Equlibrium As, mmolAs L-1
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
As
adso
rbed
, mol
As m
olA
l+Fe
-1
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
0.18
0:1 Al:Fe2:8 Al:Fe5:5 Al:Fe1:0 Al:Fe
Figure 31. Adsorption isotherms of AsIII at pH 8 on precipitated products of various Al:Fe molar ratio.
68
Equilibrium As, mmolAs L-1
0.000 0.002 0.004 0.006 0.008 0.010 0.012 0.014 0.016 0.018
As
adso
rbed
, mol
As m
olA
l+Fe
-1
0.00
0.01
0.02
0.03
0.04
0.05
0:1 Al:Fe2:8 Al:Fe5:5 Al:Fe1:0 Al:Fe
Figure 32. Adsorption isotherms of AsIII at low equilibrium AsIII concentrations at pH 5 on precipitated products of various Al:Fe molar ratio.
69
Equibrium As, mmolAs L-1
0.000 0.005 0.010 0.015 0.020 0.025 0.030
As
adso
rbed
, mol
As m
olA
l+Fe
-1
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0:1 Al:Fe2:8 Al:Fe5:5 Al:Fe1:0 Al:Fe
Figure 33. Adsorption isotherms of AsIII at low equilibrium AsIII concentrations at pH 8 on precipitated products of various Al:Fe molar ratio.
70
hydroxide is considerably weaker than that of AsV because the predominant mode of
bonding of AsIII on gibbsite and amorphous Al hydroxide is outer-sphere complexation,
compared to the predominant inner-sphere bidentate-binuclear bonding of AsV on
gibbsite.
Effect of pH
In general, more AsV was adsorbed at pH 5 than at pH 8, while higher quantities
of AsIII were adsorbed at pH 8 than at pH 5 (Figure 26 versus Figure 27 and Figure 30
versus Figure 31). Previously published data has also indicated greater adsorption of AsV
at lower pH and AsIII at higher pH (Raven et al., 1998). At pH 5, AsV exists mainly as
H2AsO4- and AsIII as neutral H3AsO3º, while at pH 8, AsV exists as HAsO4
2- with a small
fraction of H2AsO4-, and AsIII exists as neutral H3AsO3º with a small fraction of H2AsO3
-.
The net charge of the hydroxide surface is positive at pH 5; however, the proportion of
negative charge sites increase as pH increases. The higher affinity of AsV at pH 5 is
predominately due to the impact of electrostatic attraction between the negatively
charged AsV and the positively charged hydroxide surface on the overall ligand-exchange
reaction. The higher affinity of AsIII at pH 8 than at pH 5 might be attributable to the
charge characteristics of AsIII (pKa1=9.22). Maximum adsorption or inflections in
adsorption envelopes are often observed at or around the pKa of the oxyanion (Mott,
1981).
71
Arsenic Adsorption Models
Langmuir Adsorption Model
A set of Langmuir adsorption isotherms was constructed to illustrate the effect of
change in variables (b and KL) on Langmuir plots. Figure 34 illustrates three adsorption
isotherms with identical adsorption maxima (b), but with varying binding strengths (KL)
as summarized in Table 5. Aqueous concentration at which adsorption reached a
maximum increased as the KL value was increased. When b is fixed, the slopes (1/b) of
the linearly transformed Langmuir function, C/q = (1 / KLb) + (C / b) (Equation [2]), are
identical; however, the intercept (1/KLb) of the function increases as the aqueous
concentration at which the adsorption maximum is achieved decreases (Figure 35).
Figure 36 illustrates three adsorption isotherms with identical intercepts (1/KLb)
of the linearly transformed Langmuir function and increasing adsorption maxima as
summarized in Table 5. Because 1/KLb was fixed, binding energy (KL) decreased as
adsorption maximum (b) was increased. The linearly transformed Langmuir functions
exhibit an increase in slope (1/b) as adsorption maxima (b) was decreased (Figure 37).
Figure 38 illustrates three hypothetical adsorption isotherms at fixed binding
strength (KL) with varying adsorption maxima (b), as summarized in Table 5. As
adsorption maxima (b) was increased both the slope (1/b) and intercept (1/KLb) of the
linearly transformed Langmuir equation decreased (Figure 39).
72
Table 5. Calculated adsorption maxima (b), binding constants (KL), and linear Langmuir functions of three sets of hypothetical data.
Adsorbent b KL Linear equationA 0.1 500 Y=10X+0.02B 0.1 250 Y=10X+0.04C 0.1 167 Y=10X+0.06D 0.0833 600 Y=12X+0.02E 0.0714 700 Y=14X+0.02F 0.05 1000 Y=20X+0.02G 0.1 500 Y=10X+0.02H 0.08 500 Y=12.5X+0.025I 0.05 500 Y=20X+0.04
73
C
0.00 0.02 0.04 0.06 0.08 0.10
q
0.00
0.02
0.04
0.06
0.08
0.10
ABC
Figure 34. Langmuir adsorption isotherms for three sets of hypothetical data with fixed adsorption maxima (b) and varying bonding strength (KL) as summarized in Table 5.
74
C
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07
C/q
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
A --- Y=10X+0.02B --- Y=10X+0.04C --- Y=10X+0.06
0.000 0.002 0.0040.00
0.02
0.04
Figure 35. Linear regression using the linear form of the Langmuir equation for three sets of hypothetical data with fixed adsorption maxima (b) and varying bond strength (KL), as summarized in Table 5.
75
C
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
q
0.00
0.02
0.04
0.06
0.08
0.10
DEF
Figure 36. Langmuir adsorption isotherms of three sets of hypothetical data with fixed intercept of the linear Langmuir equation (1/KLb) and increasing adsorption maxima (b), as summarized in Table 5.
76
C
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
C/q
0.0
0.5
1.0
1.5
2.0
2.5
3.0
D --- Y=12X+0.02E --- Y=14X+0.02F --- Y=20X+0.02
Figure 37. Linear regression using the linear form of Langmuir equation for three sets of hypothetical data with fixed intercept (1/KLb) and increasing adsorption maxima (b), as summarized in Table 5.
77
igure 38. Langmuir adsorption isotherms of three sets of hypothetical data with fixed
C
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
q
0.00
0.02
0.04
0.06
0.08
0.10
0.12
GHI
Fbinding strength (KL) and varying adsorption maxima (b), as summarized in Table 5.
78
C
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
C/q
0.0
0.5
1.0
1.5
2.0
2.5
3.0
G --- Y=10X+0.02H --- Y=12.5X+0.025 I --- Y=20X+0.04
0.000 0.004 0.008 0.0120.0
0.1
0.2
0.3
Figure 39. Linear regression using the linear form of the Langmuir equation for three sets f hypothetical data with fixed binding strength (KL) and varying adsorption maxima (b), s summarized in Table 5.
oa
79
The Langmuir parameters for the adsorption isotherms of AsV and AsIII on the
l:Fe hydroxides are summarized in Table 6 and Table 7. In all cases, the linear
Langm
8 for
d
ere
s
ior (Figure 30 and Figure 31). Adsorption of AsIII
as mo
ons.
as Al:Fe
A
uir model described the data considerably better for AsV than for AsIII. The b
values indicate that the adsorption maxima of AsV were higher at pH 5 than at pH
each of the Al:Fe molar ratios. Although the adsorption behavior of AsV on the 0:1 an
2:8 Al:Fe hydroxides were similar at both pH 5 and 8 (Figure 26 and Figure 27), the KL
values of the Langmuir adsorption isotherms indicate that AsV was more strongly
adsorbed to the 0:1 Al:Fe hydroxide than the 2:8 Al:Fe hydroxide (Table 6, Figure 40,
and Figure 41). The KL values for adsorption of AsV on the 0:1 Al:Fe hydroxide w
23,000 at pH 5 and 14,000 at pH 8,while the KL values for adsorption of AsV on the 2:8
Al:Fe hydroxide were 16,000 at pH 5 and 9,700 at pH 8 (Table 6). Also, the b values
indicate slightly higher adsorption maximum with the 0:1 Al:Fe hydroxide compared to
that of the 2:8 Al:Fe hydroxide.
In the case of AsIII, the b and KL values were higher at pH 8 than at pH 5, which i
comparable with its adsorption behav
w re highly affected by Al:Fe molar ratio than in the case of AsV (Figure 42 and
Figure 43). The KL values for adsorption of AsIII on the 0:1 Al:Fe hydroxide were 288 at
pH 5 and 565 at pH 8, while the KL values for adsorption of AsIII on the 2:8 Al:Fe
hydroxide were 159 at pH 5 and 200 at pH 8 (Table 7). The decrease in adsorption
maximum as affected by Al:Fe ratio was more severe with AsIII than with AsV.
AsV exhibited higher KL and b values than those of AsIII under similar conditi
The adsorption maxima and bonding strengths of both of AsV and AsIII increased
80
Table 6. Coefficients of determination (r2), calculated adsorption maxima (b), and inding constants (KL) of AsV adsorption as affected by Al:Fe molar ratios derived by angmuir linear functions.
inding constants (KL) of AsIII adsorption as affected by Al:Fe molar ratios derived by angmuir linear functions.
Adsorbent pH r2 b KL
8
8 0.9989 0.0858 97135:5 Al:Fe 5 0.9971 0.0846 6219
8 0.9973 0.0651 46521:0 Al:Fe 5 0.9971 0.0505 1376
8 0.9957 0.0358 910
bL
0:1 Al:Fe 5 0.9985 0.1073 23298 0.9996 0.0876 14266
2:8 Al:Fe 5 0.9968 0.1062 15691
Table 7. Coefficients of determination (r2), calculated adsorption maxima (b), and bL
Adsorbent pH r2 b KL
0:1 Al:Fe 5 0.9574 0.0812 2888 0.9824 0.0843 565
0.9168 0.0377 1048 0.8828 0.0598 145
2:8 Al:Fe 5 0.946 0.0577 1598 0.905 0.0785 200
5:5 Al:Fe 5
81
C, mmolAs L-1
0.000 0.005 0.010 0.015 0.020 0.025
C/q
, mm
olA
l+Fe
L-1
0.0
0.1
0.2
0.3
0.4
0.5
0:1 Al:Fe --- Y=9.32X+0.00042:8 Al:Fe --- Y=9.41X+0.00065:5 Al:Fe --- Y=11.82X+0.00191:0 Al:Fe --- Y=19.81X+0.0144
Figure 40. Linear regression using the linear form of the Langmuir equation for valuation of adsorption of AsV by Al:Fe hydroxides at pH 5, as summarized in Table 6. e
82
C, mmolAs L-1
0.000 0.005 0.010 0.015 0.020 0.025 0.030
C/q
, mm
olA
l+Fe
L-1
0.0
0.2
0.4
0.6
0.8
0:1 Al:Fe --- Y=11.41X+0.00082:8 Al:Fe --- Y=11.66X+0.00125:5 Al:Fe --- Y=15.35X+0.00331:0 Al:Fe --- Y=27.93X+0.0307
Figure 41. Linear regression using the linear form of the Langmuir equation for evaluation of adsorption of AsV by Al:Fe hydroxides at pH 8, as summarized in Table 6.
83
C, mmolAs L-1
0.00 0.01 0.02 0.03 0.04 0.05
C/q
, mm
olA
l+Fe
L-1
0.0
0.2
0.4
0.6
0.8
1.0
0:1 Al:Fe --- Y=12.31X+0.04272:8 Al:Fe --- Y=17.33X+0.1095:5 Al:Fe --- Y=26.54X+0.256
Figure 42. Linear regression using the linear form of the Langmuir equation for evaluation of adsorption of AsIII by Al:Fe hydroxides at pH 5, as summarized in Table 7.
84
C, mmolAs L-1
0.000 0.005 0.010 0.015 0.020 0.025
C/q
, mm
olA
l+Fe
L-1
0.0
0.1
0.2
0.3
0.4
0.5
0:1 Al:Fe --- Y=11.87X+0.02102:8 Al:Fe --- Y=12.74X+0.06385:5 Al:Fe --- Y=16.73X+0.116
Figure 43. Linear regression using the linear form of the Langmuir equation for evaluation of adsorption of AsIII by Al:Fe hydroxides at pH 8, as summarized in Table 7.
85
molar ratio decreased; however, this trend was more noticeable in the case of AsIII. In
general, these trends indicate the higher affinity of AsV on both Fe hydroxide and Al
hydroxide compared to that of AsIII, and higher retention of both AsV and AsIII on Fe
hydroxide compared to Al hydroxide.
Freundlich Adsorption Model
Sets of hypothetical Freundlich adsorption isotherms were constructed to illustrate
the effects of the variables, N and KF, on the Freundlich isotherm, q = KFCN (Equation
[3]), and its linear transformation, log q = N log C + log KF (Equation [4]) with 0 ≤ N ≤
1. Figure 44 illustrates three hypothetical adsorption isotherms with identical KF but
varying N, as summarized in Table 8. When N is 1 (Figure 44 – C), the C-curve isotherm
is obtained (Essington, 2004), that is, the slope of the Freundlich isotherm remains
constant regardless of the surface coverage. As N is decreased, within the constraints of
0 ≤ N ≤ 1, the initial slope of the adsorption isotherm is greater. N is most strongly
influenced by the initial slope of the adsorption isotherm. A very small N value is
indicative of quantitative adsorption at low C. The slope of the linearly transformed
Freundlich equation increases with an increase in the N value, since N is the slope of the
linear Freundlich function (Figure 45). Because KF values were fixed in this example,
the intercepts (log KF) of the linear Freundlich function are identical.
86
Table 8. Calculated N and KF of Freundlich linear functions of two sets of hypothetical data.
Adsorbent N KF Linear equationA 0.1 0.3162 Y=0.1X-0.2B 0.5 0.3162 Y=0.5X-0.2C 1 0.3162 Y=X-0.2D 0.5 0.631 Y=0.5X-0.2E 0.5 0.3162 Y=0.5X-0.5F 0.5 0.1 Y=0.5X-1
87
C
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
q
0.00
0.05
0.10
0.15
0.20
0.25
0.30
ABC
Figure 44. Freundlich adsorption isotherms, q = KFCN, for three sets of hypothetical data, with fixed KF but with varying N, as summarized in Table 8.
88
log C
-5 -4 -3 -2 -1 0
log
q
-5
-4
-3
-2
-1
0
A --- Y=0.1X-0.2B --- Y=0.5X-0.2C --- Y=X-0.2
Figure 45. Linear regression lines of the Freundlich linear equation, log q = N log C + log
KF, for three sets of hypothetical data from Figure 44, with fixed KF and varying N. The regression equations for these lines are summarized in Table 8.
89
Figure 46 illustrates three hypothetical adsorption isotherms with identical N but
varying KF, as summarized in Table 8. A higher KF value also indicates a greater
adsorption when N and C are fixed, as it is apparent from its mathematical relationship, q
= KFCN. Linearly transformed Freundlich equations exhibit parallel lines as they have
identical slopes, N (Figure 47). The intercepts, logKF, of these functions increase as the
KF value increases.
The Freundlich parameters, N and KF, for the adsorption of AsV and AsIII on the
Al:Fe hydroxides are summarized in Table 9 and Table 10. With both AsV and AsIII
adsorption isotherms, calculated N values were not appreciably different at pH 5 than at
pH 8, and were also not significantly affected by Al:Fe molar ratio (Table 9 and Table
10). The N values of AsV adsorption isotherms were similar, approximately 0.12 to 0.16;
therefore, nearly parallel lines of the linearly transformed Freundlich lines were observed
(Figure 48 and Figure 49). The N values of AsIII adsorption isotherms were also similar,
approximately 0.41 to 0.50, and approximately parallel lines of the linearly transformed
Freundlich lines were also observed (Figure 50 and Figure 51). The similar values of N
within the set of AsV adsorption isotherms and within the set of AsIII adsorption isotherms
indicate that the initial increase in adsorption followed similar trends at the various Al:Fe
molar ratios, for both AsIII and AsV.
The N values of AsV adsorption isotherms (0.12 ≤ N ≤ 0.16) were considerably
smaller than those of AsIII (0.4137 ≤ N ≤ 0.497). This trend indicates the greater slopes
of adsorption isotherms at low C in the case of AsV compared to AsIII, which indicates a
higher affinity of AsV than AsIII to the hydroxide surface. When AsV was quantitatively
adsorbed at the lower As:(Al+Fe) molar ratios, the equilibrium concentration (C) was
90
C
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
q
0.00
0.05
0.10
0.15
0.20
0.25
DEF
Figure 46. Freundlich adsorption isotherms, q=KFCN, for three sets of hypothetical data, with fixed N but with varying KF, as summarized in Table 8.
91
log C
-4.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5
log
q
-3.5
-3.0
-2.5
-2.0
-1.5
-1.0
-0.5
D --- Y=0.5X-0.2E --- Y=0.5X-0.5F --- Y=0.5X-1
Figure 47. Linear regression lines of the Freundlich linear equation, log q = N log C + log KF, for three sets of hypothetical data from Figure 46, with fixed N and varying KF, as summarized in Table 8.
92
Table 9. Coefficients of determination (r2), and calculated N and KF values for AsV adsorption by Al:Fe hydroxides, as derived using the Freundlich linear functions.
able 10. Coefficients of determination (r2), and calculated N and KF values for AsIII
Adsorbent pH r2 N KF
0:1 Al:Fe 5 0.9928 0.1433 0.23688 0.9869 0.1226 0.1586
2:8 Al:Fe 5 0.9984 0.155 0.24438 0.9894 0.116 0.1463
5:5 Al:Fe 5 0.9983 0.1403 0.16048 0.9817 0.1372 0.1043
1:0 Al:Fe 5 0.9788 0.1586 0.09298 0.9727 0.161 0.0638
Tadsorption by Al:Fe hydroxides, as derived using the Freundlich linear functions.
Adsorbent pH r2 N KF
0:1 Al:Fe 5 0.9921 0.4724 0.51198 0.9952 0.4137 0.4907
2:8 Al:Fe 5 0.9895 0.4859 0.30378 0.9908 0.4966 0.4823
5:5 Al:Fe 5 0.9696 0.4867 0.16588 0.9846 0.4795 0.2934
93
log C, mmolAs L-1
-5 -4 -3 -2 -1
log
q, A
s ad
sorb
ed m
olA
s m
olA
l+Fe
-1
-1.5
-1.4
-1.3
-1.2
-1.1
-1.0
-0.9 0:1 Al:Fe --- Y=0.1433X-0.62562:8 Al:Fe --- Y=0.1550X-0.61205:5 Al:Fe --- Y=0.1430X-0.79471:0 Al:Fe --- Y=0.1586X-1.032
Figure 48. Linear regression lines of the linearly transformed Freundlich adsorption
otherms for adsorption of AsV by various Al:Fe hydroxides at pH 5, as summarized in able 9.
isT
94
log C, mmolAs L-1
-4.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0
log
q, A
s ad
sorb
ed m
olA
s mol
Al+
Fe-1
-1.6
-1.5
-1.4
-1.3
-1.2
-1.1
-1.0 0:1 Al:Fe --- Y=0.1226X-0.79962:8 Al:Fe --- Y=0.1160X-0.83475:5 Al:Fe --- Y=0.1372X-0.91611:0 Al:Fe --- Y=0.1610X-1.195
Figure 49. Linear regression lines of the linearly transformed Freundlich adsorption otherms for adsorption of AsV by various Al:Fe hydroxides at pH 8, as summarized in able 9.
isT
95
log C, mmolAs L-1
-4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0
log
q, A
s ad
sorb
ed m
olA
s mol
Al+
Fe-1
-2.4
-2.2
-2.0
-1.8
-1.6
-1.4
-1.2
-1.0
0:1 Al:Fe --- Y=0.4724X-0.30922:8 Al:Fe --- Y=0.4859X-0.51755:5 Al:Fe --- Y=0.4867X-0.7805
Figure 50. Linear regression lines of the linearly transformed Freundlich adsorption isotherms for adsorption of AsIII by various Al:Fe hydroxides at pH 5, as summarized in
able 10. T
96
log C, molAs L-1
-4.0 -3.5 -3.0 -2.5 -2.0 -1.5
log
q, A
s ad
sorb
ed m
olA
s mol
Al+
Fe-1
-2.2
-2.0
-1.8
-1.6
-1.4
-1.2
-1.0
0:1 Al:Fe --- Y=0.4137X-0.30922:8 Al:Fe --- Y=0.4966X-0.31625:5 Al:Fe --- Y=0.4795X-0.5326
Figure 51. Linear regression lines of the linearly transformed Freundlich adsorption otherms for adsorption of AsIII by various Al:Fe hydroxides at pH 8 as summarized in able 10.
isT
97
zero. In determination of the linearly transformed Freundlich equation, log q = N log C +
ller
nces in N values within AsV or AsIII were small, the
differen N. In
at
rms with
the 0:1
t
ed relatively well by the
Freund
log KF (Equation [3]), the data points in which the AsV was quantitatively adsorbed could
not be considered, since log of zero is undefined. Because the points where AsV was
quantitatively adsorbed were eliminated, the actual values of N are expected to be sma
than the calculated values.
Because the differe
ces between adsorption isotherms were mainly reflected by KF rather than
general, there was a trend of higher KF values with lower Al:Fe molar ratio with both AsV
and AsIII; however, K values were higher at pH 5 as opposed to pH 8 in case of AsFV, and
those of AsIII were higher at pH 8 rather than pH 5. These trends indicate the greater
adsorption of AsV at pH 5 compared to that at pH 8, and the greater adsorption of AsIII
pH 8 as opposed to pH 5. The interpretation of the Freundlich parameters is compatible
with the actual isotherms (Figure 26, Figure 27, Figure 30, and Figure 31).
The linearly transformed Freundlich functions of AsV adsorption isothe
and 2:8 Al:Fe hydroxides nearly overlapped regardless of pH (Figure 48 and
Figure 49), whereas, those of the AsIII adsorption isotherms had significantly differen
intercepts with similar slopes (Figure 50, and Figure 51). This observation, along with
the adsorption isotherms and the results of the Langmuir model indicate that AsV
adsorption on the 0:1 and 2:8 Al:Fe hydroxides is comparable.
Adsorption isotherms of both AsV and AsIII were describ
lich model, since the correlation coefficients (r2) were always greater than 0.97
(Table 9 and Table 10).
98
The Langmuir Model versus the Freundlich Model
Higher coef Langmuir model
than wi
odel
tion Envelopes
Adsorption of AsV fo ith respect to pH, for the
various
52).
f
containing hydroxides, due to the presence of the crystalline hydroxides, bayerite and
ficients of determination (r2) were observed with the
th the Freundlich model when they were used to describe the adsorption isotherms
of AsV on the various hydroxides (Table 6 versus Table 9). However, higher coefficients
of determination (r2) were observed with the Freundlich model rather than the Langmuir
model for the AsIII adsorption isotherms (Table 7 and Table 10). In general, the
Langmuir model better described AsV adsorption isotherms, and the Freundlich m
better described AsIII adsorption isotherms.
Arsenic Adsorp
llowed the same general trends, w
Al:Fe hydroxides under the experimental conditions [As:(Al+Fe) = 0.05:1]
utilized for the adsorption envelopes; however, adsorption of AsV on the 1:0 Al:Fe
hydroxide was much less compared to that of the Fe-containing hydroxides (Figure
AsV was quantitatively adsorbed on 0:1, 2:8, and 5:5 Al:Fe hydroxides at pH 3 to 6.5.
Electrostatic attraction between the negatively charged AsV and the positively charged
hydroxide surface impacts the high affinity for AsV within this pH range. Adsorption o
AsV decreased gradually with decreasing pH at pH > 7, as the repulsive potential between
AsV and the hydroxide surface and the competition of OH- for surface-adsorption sites
increased. The negative charge of AsV increases at pH > 6 (Figure 2), because the pKa2
of As is 6.97. As was never quantitatively adsorbed by the 1:0 Al:Fe hydroxide, and
an adsorption maximum was observed at pH 5.2, with 76 % of total As adsorbed. The
1:0 Al:Fe hydroxide has a low concentration of adsorption sites compared to the Fe-
V V
V
99
pH
3 4 5 6 7 8 9 10 11
As
adso
rbed
, %
0
20
40
60
80
100
0:1 Al:Fe2:8 Al:Fe5:5 Al:Fe1:0 Al:Fe
Figure 52. Adsorption envelopes of AsV in 0.1 M NaCl at an As:(Al+Fe) molar ratio of 0.05:1, at various Al:Fe molar ratios.
100
gibbsite, which were identified by XRD and TEM as major constituents of the
precipitated hydroxide. A similar trend in AsV adsorption by amorphous Al hydroxide as
vior
of the
s
5.2.
he
bsite
0:1, 2:8, and 5:5 Al:Fe hydroxides resulted in similar shapes of
adsorp
a function of pH was observed by Anderson et al. (1976). This adsorption beha
might be influenced by the solubility of Al hydroxide. The 0:1 Al:Fe hydroxide is
subject to enhanced dissolution at low pH (i.e. pH < 4.5; Lindsay, 1979), which
contributes to the solubility of AsV (Figure 53). Though pzc of the Al hydroxide is
typically around pH 9 (Hsu, 1989), a specifically adsorbed anion can shift the pzc
hydroxide surface to lower pH values, making the surface charge at a given pH more
negative (Mott, 1981). Anderson et al. (1976) observed that the pH of the isoelectric
point decreased from 8.5 to 4.6 as increasing amounts of AsV were added to amorphou
Al hydroxide. This trend explains the sharp decrease in AsV adsorption starting at pH
Electrostatic repulsion between AsV and the Al hydroxide increased because of the
adsorbed AsV, and further adsorption was reduced. The pzc of Fe-containing hydroxides
would also have been lowered due to the adsorbed AsV(Jain et al., 1999); however, t
sharp decrease in adsorption was not observed because of the significantly higher
concentration of adsorption sites per unit weight of adsorbent with the poorly crystalline
Fe-containing hydroxides compared to the Al hydroxide that was dominated by gib
and bayerite (Figure 7).
Unlike AsV adsorption, AsIII adsorption decreased substantially as Al:Fe molar
ratio increased; however,
tion envelopes (Figure 54). Adsorption maxima of AsIII were observed in the pH
range of 8 to 9 with the 0:1, 2:8, and 5:5 Al:Fe hydroxides. AsIII adsorption by the 1:0
Al:Fe hydroxide was negligible across the entire pH range of 3 to 11. At the adsorption
101
Figure 53. The hydrolysis species of Al3+ ion in equilibrium with gibbsite as a function of pH, calculated using thermodynamic constants tabulated in Lindsay (1979).
pH
3 4 5 6 7 8 9 10
log
activ
ity
-20
-15
-10
-5
0
AlOH2+
Al(OH)2+
Al(OH)3
O
Al(OH)4-
Al(OH)52-
Al2(OH)24+
Total Al
102
pH
3 4 5 6 7 8 9 10 11
As
adso
rbed
, %
0
20
40
60
80
100
0:1 Al:Fe2:8 Al:Fe5:5 Al:Fe1:0 Al:Fe
Figure 54. Adsorption envelopes of AsIII in 0.1 M NaCl at an As:(Al+Fe) molar ratio of 0.05:1, at various Al:Fe molar ratios.
103
maxima, the 0:1, 2:8, and 5:5 Al:Fe hydroxides adsorbed 92, 85, and 73 %, respectively,
f the initially added AsIII. Adsorption of AsIII decreased above pH 9 because the
easing
of
Al content due to the negligible AsIII adsorption by Al
hydrox
us
was not strongly influenced by Al substitution except with the 1:0 Al:Fe
o
negatively charged species dominate above pH 9.22 (pKa1), and the repulsive forces
between AsIII and the hydroxide surfaces are substantial and will increase with incr
pH. The decrease in AsIII adsorption with Al substitution indicates the higher affinity
AsIII by Fe3+ relative to Al3+. This result is supported by the result of a previous study by
Ferguson and Anderson (1974), in which it was also observed that AsIII was not readily
adsorbed by Al hydroxide.
Although the decrease in AsIII adsorption was expected to occur in the same
proportion as the increase in
ide, the AsIII adsorption maximum decreased by approximately 4 % with 10 %
increase in Al content on average. This phenomenon might be due to a heterogeneo
distribution of Al within the structure of the hydroxide. There is also a possibility that
Fe3+ might be preferentially residing at the outer layer of the aggregates; although, Fe
would be expected to hydrolyze first because the log Kº of the first Fe3+ hydrolysis is –
2.19 and that of the first Al hydrolysis is –5.02 (Lindsay, 1979). It is also possible that
Al3+ was able to adsorb some AsIII, which might indicate a difference in the affinity of
AsIII to pure Al hydroxide surface sites compared to Al3+ sties at the Al:Fe hydroxide
surface, i.e., the presence of structure Fe3+ might have influenced AsIII adsorption at Al3+
surface sites.
Under the conditions of this experiment [As:(Al+Fe) = 0.05:1 in 0.1 M NaCl],
AsV adsorption
104
hydrox
with increase in Al substitution. However,
AsV ca
cted by Counterion
AsV was ge xides in larger
amounts in the pr H 5, AsV was
quantit
of
lar ratio
and the surface-charge characteristics of the Al:Fe hydroxide. At pH 5,
the attr
and the
ides. Further work is needed to more fully evaluate the influence of pH on AsV
adsorption at high As:(Al+Fe) molar ratios.
In the water treatment scenario, AsIII removal by Al-substituted hydroxide might
be problematic since its adsorption decreased
n be removed by Al-substituted hydroxides as efficiently as by pure Fe hydroxide
when a sufficient amount of hydroxide is present.
Adsorption of Arsenic as Affected by Counterion
Adsorption Isotherms of AsV as Affe
nerally adsorbed on both the 0:1 and 2:8 Al:Fe hydro
esence of Ca2+ than Na+ (Figure 55 and Figure 56). At p
atively adsorbed by both 0:1 and 2:8 Al:Fe hydroxides up to approximately 0.10
As:(Al+Fe) molar ratio (Figure 55). Slightly higher retention of AsV was observed in the
presence of Ca2+ compared to Na+ as As:(Al+Fe) molar ratio was increased. At pH 8,
AsV was adsorbed quantitatively by both 0:1 and 2:8 Al:Fe hydroxides up to
approximately 0.06 As:(Al+Fe) molar ratio (Figure 56). Considerably higher retention
AsV was observed in the presence of Ca2+ compared to Na+ as As:(Al+Fe) mo
was increased.
The differences in AsV retention due to pH can be explained by the electrostatic
attraction of AsV
action between the positively charged Al:Fe hydroxide surface and the negatively
charged AsV species is so strong that AsV adsorption was favored regardless of
counterion. At pH 8, the Al:Fe hydroxide surface is negatively charged as a result of
both the pH dependent negative charge character of the variable charge mineral
105
Equilibrium As, mmolAs L-1
0.00 0.02 0.04 0.06 0.08 0.10
As
adso
rbed
, mol
As m
olA
l+Fe
-1
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
0.18
0.20
0.22
Ca, 0:1 Al:FeCa, 2:8 Al:FeNa, 0;1 Al:FeNa, 2:8 Al:Fe
Figure 55. Adsorption isotherms of AsV in 0.1 molCa L-1 and 0.1 molNa L-1 at pH 5 as
affected by counterion and Al:Fe molar ratio.
106
Equilibrium As, mmolAs L-1
0.00 0.02 0.04 0.06 0.08 0.10
As
adso
rbed
, mol
As m
olA
l+Fe
-1
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
0.18
0.20
Ca, 0:1 Al:FeCa, 2:8 Al:FeNa, 0:1 Al:FeNa, 2:8 Al:Fe
Figure 56. Adsorption isotherms of AsV in 0.1 molCa L-1 and 0.1 molNa L-1 at pH 8 as
affected by counterion and Al:Fe molar ratio.
107
specific adsorption of anionic AsV species (Mott, 1981); therefore, further adsorption of
sV was reduced. Because of the repulsive potential between the negatively charged
ydroxide surface and the negatively charged AsV, a significant effect of divalent cation,
l
Adsorption Envelopes of As as Affected by Counterion
Ca vs. Na
The adsorption envelopes of As indicated that the adsorption maximum was in
the pH range of a le it was
he pH range of approximately 3 to 7.5 at 0.025:1 As:(Al+Fe) molar ratio,
f Al
n
A
h
Ca2+, was observed at pH 8. Enhanced retention of As by soil and Al and Fe hydroxides
in the presence of Ca2+ compared to Na+ as the counterion has been reported by severa
investigators (Parks et al., 2003; Smith et al., 2002b). The possible scenarios by which
Ca2+ could enhance the adsorption of AsV are discussed in the hypothesis section below.
Although significant differences in adsorption as affected by Al substitution were
not observed at pH 5, slightly better retention of AsV by the 0:1 Al:Fe hydroxide was
observed at pH 8 (Figure 55 and Figure 56). There is a possibility that Ca2+ can better
enhance the retention of AsV without Al substitution; however, further study is needed, as
there is also a chance of a potential error associated with the determination of adsorbed
AsV concentration at high dissolved AsV concentrations.
V
2+ +
V
pproximately 3 to 5 at 0.1:1 As:(Al+Fe) molar ratio, whi
observed in t
regardless of counterion (Figures 57 – 61). At pH above the adsorption maxima, AsV
adsorption decreased gradually as pH increased in the presence of Ca2+ regardless o
substitution and As:(Al+Fe) molar ratio, while retention of AsV decreased more rapidly
as pH was increased in the presence of Na+ (Figures 57 – 61). In summary, the retentio
of AsV was higher in the presence of Ca2+ than of Na+ with both the 0:1 and 2:8
108
Figure 57. Adsorption envelopes of AsV in 0.1 molCa L-1 and 0.1 molNa L-1 as affected by
counterion and Al substitution at 0.1:1 As:(Al+Fe) molar ratio.
pH
3 4 5 6 7 8 9 10 11
As
adso
rbed
, %
0
20
40
60
80
100
Ca, 0:1 Al:FeCa, 2:8 Al:FeNa, 0:1 Al:FeNa, 2:8 Al:Fe
109
pH
3 4 5 6 7 8 9 10 11
As
adso
rbed
, %
0
20
40
60
80
100
Ca, 0:1 Al:FeCa, 2:8 Al:FeNa, 0:1 Al:FeNa, 2:8 Al:Fe
Figure 58. Adsorption envelopes of AsV in 0.01 molCa L-1 and 0.01 molNa L-1 as affected
by counterion and Al substitution at 0.1:1 As:(Al+Fe) molar ratio.
110
pH
3 4 5 6 7 8 9 10 11
As
adso
rbed
, %
0
20
40
60
80
100
Ca, 0:1 Al:FeCa, 2:8 Al:FeNa, 0:1 Al:FeNa, 2:8 Al:Fe
Figure 59. Adsorption envelopes of AsV in 0.001 molCa L-1 and 0.001 molNa L-1as affected
by counterion and Al substitution at 0.1:1 As:(Al+Fe) molar ratio.
111
pH
3 4 5 6 7 8 9 10 11
As
adso
rbed
, %
0
20
40
60
80
100
Ca, 0:1 Al:FeCa, 2:8 Al:FeNa, 0:1 Al:FeNa, 2:8 Al:Fe
Figure 60. Adsorption envelopes of AsV in 0.1 molCa L-1 and 0.1 molNa L-1 as affected by
counterion and Al substitution at 0.025:1 As:(Al+Fe) molar ratio.
112
pH
3 4 5 6 7 8 9 10 11
As
adso
rbed
, %
0
20
40
60
80
100
Ca, 0:1 Al:FeCa, 2:8 Al:FeNa, 0:1 Al:FeNa, 2:8 Al:Fe
Figure 61. Adsorption envelopes of AsV in 0.01 molCa L-1 and 0.01 molNa L-1 as affected
by counterion and Al substitution at 0.025:1 As:(Al+Fe) molar ratio.
113
hydroxides, especially at pH>7; however, differences in adsorption maxima were not
rption envelopes were observed with 0.1, 0.01, and
0.001 m
re
ption
f
erg and
titution
al, adsorption envelopes of the 2:8 Al:Fe hydroxide overlapped with those
2:8
strongly influenced by either counterion or its concentration.
Counterion Concentration Effect
Similar trends in AsV adso
olCa L-1 for both 0:1 and 2:8 Al:Fe hydroxides and at both 0.025:1 and 0.1:1
As:(Al+Fe) molar ratios (Figures 62 – 65). Also, similar trends in AsV adsorption we
observed in 0.1, 0.01, and 0.001 molNa L-1 with both hydroxides and at both the 0.025:1
and 0.1:1 As:(Al+Fe) molar ratios (Figures 66 – 69). Though AsV adsorption was
strongly influenced by counterion, i.e., Ca2+ vs. Na+, as discussed previously, adsor
was not strongly affected by counterion concentration at either the 0.025:1 or 0.1:1
As:(Al+Fe) molar ratio regardless of pH and Al:Fe molar ratio. The independence o
AsV adsorption on counterion concentration indicates inner-sphere complexation as the
AsV retention mechanism (McBride, 1997). The similar trends with both 0:1 and 2:8
hydroxides indicate the similar modes of bonding of AsV in these two systems.
Independence of AsV adsorption from ionic strength was also observed by Goldb
Johnson (2001), who studied AsV adsorption on amorphous Fe oxides in 0.01 to 1.0 M
NaCl.
Al subs
In gener
of the 0:1 hydroxide regardless of experimental variables, which include counterion,
counterion concentration, and pH (Figure 57 – 61). Adsorption of AsV by the 0:1 and
Al:Fe hydroxides was similarly enhanced in the presence of Ca2+. Improved retention of
114
pH
3 4 5 6 7 8 9 10 11
As
adso
rbed
, %
0
20
40
60
80
100
0.1 molCa L-1
0.01 molCa L-1
0.001 molCa L-1
Figure 62. Adsorption envelopes of AsV with the 0:1 Al:Fe hydroxide at 0.1:1 As:(Al+Fe) molar ratio as affected by Ca counterion concentration.
115
pH
3 4 5 6 7 8 9 10 11
As
adso
rbed
, %
0
20
40
60
80
100
0.1 molCa L-1
0.01 molCa L-1
0.001 molCa L-1
Figure 63. Adsorption envelopes of AsV with the 2:8 Al:Fe hydroxide at 0.1:1 As:(Al+Fe) molar ratio as affected by Ca counterion concentration.
116
pH
3 4 5 6 7 8 9 10 11
As
adso
rbed
, %
0
20
40
60
80
100
0.1 molCa L-1
0.01 molCa L-1
Figure 64. Adsorption envelopes of AsV with the 0:1 Al:Fe hydroxide at 0.025:1 As:(Fe+Al) molar ratio as affected by Ca counterion concentration.
117
pH
3 4 5 6 7 8 9 10 11
As
adso
rbed
, %
0
20
40
60
80
100
0.1 molCa L-1
0.01 molCa L-1
Figure 65. Adsorption envelopes of AsV with the 2:8 Al:Fe hydroxide at 0.025:1 As:(Fe+Al) molar ratio as affected by Ca counterion concentration.
118
pH
3 4 5 6 7 8 9 10 11
As
adso
rbed
, %
0
20
40
60
80
100
0.1 molNa L-1
0.01 molNa L-1
0.001 molNa L-1
Figure 66. Adsorption envelopes of AsV with the 0:1 Al:Fe hydroxide at 0.1:1 As:(Fe+Al) molar ratio as affected by Na counterion concentration.
119
pH
3 4 5 6 7 8 9 10 11
As
adso
rbed
, %
0
20
40
60
80
100
0.1 molNa L-1
0.01 molNa L-1
0.001 molNa L-1
Figure 67. Adsorption envelopes of AsV with the 2:8 Al:Fe hydroxide at 0.1:1 As:(Fe+Al) molar ratio as affected by Na counterion concentration.
120
pH
3 4 5 6 7 8 9 10 11
As
adso
rbed
, %
0
20
40
60
80
100
0.1 molNa L-1
0.01 molNa L-1
0.001 molNa L-1
Figure 68. Adsorption envelopes of AsV with the 0:1 Al:Fe hydroxide at 0.025:1 As:(Fe+Al) molar ratio as affected by Na counterion concentration.
121
igure 69. Adsorption envelopes of AsV with the 2:8 Al:Fe hydroxide at 0.025:1
pH
3 4 5 6 7 8 9 10 11
As
adso
rbed
, %
0
20
40
60
80
100
0.1 molNa L-1
0.01 molNa L-1
0.001 molNa L-1
FAs:(Fe+Al) molar ratio as affected by Na counterion concentration.
122
AsV by both Fe and Al hydroxides in the presence of Ca2+ was also reported by
(Goldberg and Johnson, 2001).
Hypotheses
There are three possible scenarios by which Ca2+ could enhance the adsorption of
AsV: (i) precipitation of Ca arsenate, (ii) cation bridging by Ca , and (iii) reduced
repulsive potentials in the presence of counterion Ca . The formation of Ca arsenate has
been reported at high pH (pH > 7.3) (Bothe and Brown, 1999). Jing et al. (2003)
observed the reduced mobility of As from cement treated Fe sludge at pH 11.32, due to
the formation of Ca arsenate. The formation of Ca arsenate has pH and Ca/As molar
ratio requirements as summarized in Table 3 (Bothe and Brown, 1999). For example, the
Ca/As molar ratio needs to be 1.5 –1.67 and pH needs to be in the range of 7.32 – 11.18
to form Ca3(AsO4)2•32/3H2O or Ca3(AsO4)2• 41/4H2O (Bothe and Brown, 1999). Other
forms of Ca arsenate require even higher pH (pH > 9.77). In the current experiment, the
differences in As retention due counterion were observed to start at approximately pH 6.
When the net charge of the hydroxide surface is negative, Ca could possibly
function as a bridge for the adsorption of negatively charged As (Figure 70). This
cation bridging could possibly enhance the retention of As ; however, Parks et al. (2003)
eliminated this possibility by diffuse layer modeling. They concluded that the thinner
diffuse double layer formed by Ca compared to Na minimizes the repulsive potential
between the negatively charged hydroxide surface and the negatively charged As ;
therefore, Ca improves the retention of As on Al and Fe hydroxides (Figure 71).
Spectroscopic evidence will be required to verify the mechanism of enhanced As
retention in the presence of Ca .
2+
2+
V
2+
V
V
2+ +
V
2+ V
V
2+
123
-
-
-
-
-
-
-
-
Ca2+ HAsO42-Ca2+ HAsO42-
-
-
-
Ca2+ H2AsO4-
-
-
-
H2AsO4-Ca2+
Figure 70. Cation bridging by Ca2+.
Na Ca
Electrical potential
Distance from surface
Figure 71. The diffuse double layers of Ca2+ vs. Na+.
124
Desorption Envelopes of Arsenic
Desorption behavior of AsV and AsIII with phosphate as the desorbing ion was
studied at 375 : 0.05 : 1 P:As:(Al+Fe) molar ratio. Desorption of AsIII from the 1:0 Al:Fe
hydroxide was not studied because AsIII was not adsorbed by the 1:0 Al:Fe hydroxides at
any pH (Figure 54).
Most of the added AsV was adsorbed by the Fe-containing hydroxides during the
adsorption stage of the experiment; however, only 49.4 % of the AsV was adsorbed on the
1:0 Al:Fe hydroxide (Table 11). Desorption of AsV by mechanical agitation of the
aqueous suspension was negligible with the Fe-containing hydroxides; however, 1.2 % of
the adsorbed AsV was desorbed from the 1:0 Al:Fe hydroxide during the 24 h shaking
with deionized water (DIW). Adsorption of AsIII was never 100 %, and AsIII adsorption
increased as Al:Fe molar ratio was decreased, as expected from the previous adsorption
isotherm study. A significant amount of AsIII was desorbed during mechanical agitation
of the aqueous suspension, and AsIII desorption in DIW increased as Al:Fe molar ratio
was increased (Table 12). This trend reflects the differences in retention mechanism of
AsV and AsIII on Al and Fe hydroxides. The predominant mode of retention of AsV and
AsIII on Fe hydroxides is by formation of an inner-sphere bidentate binuclear surface
complex (Manning et al., 1998; Waychunas et al., 1993), although outer-sphere
complexation of AsIII has also been observed (Goldberg and Johnson, 2001). AsV forms
inner-sphere complexes with amorphous Al hydroxide, but AsIII forms only outer-sphere
complexes (Goldberg and Johnson, 2001).
125
Table 11. Proportion of AsV adsorbed during the 24 h adsorption reaction before phosphate desorption (A), proportion of adsorbed AsV after 24 h shaking with deionized water (DIW) (B), and proportion of AsV desorbed during 24 h shaking with deionized water.
able 12. Proportion of AsIII adsorbed during the 24 h adsorption reaction before onized
AdsorbentAl:Fe
AsV adsorbed during 24 h
adsorption (A)
AsV adsorbedafter 24 h
desorption with DIW (B)
AsV desorbed after 24 h desorption
with DIW (A) - (B)
%0:1 99.8 99.9 02:8 99.7 99.6 0.15:5 98.8 98.5 0.31:0 49.4 48.2 1.2
Tphosphate desorption (A), proportion of adsorbed AsIII after 24 h shaking with deiwater (DIW) (B), and proportion of AsIII desorbed during 24 h shaking with deionized water.
AdsorbentAl:Fe
AsIII adsorbed during 24 h
adsorption (A)
AsIII adsorbedafter 24 h
desorption with DIW (B)
AsIII desorbed after 24 h desorption
with DIW (A) - (B)
%0:1 91 88.3 2.72:8 79.5 75.8 3.75:5 66.4 57.3 9.1
126
Neither AsV nor AsIII was completely desorbed by phosphate from the Fe-
containing hydroxides, at any Al:Fe molar ratio and at any pH value within the range of 3
to 11; however, As desorption was always > 50% (Figure 72 and Figure 73).
In general, As exhibited similar desorption patterns regardless of Al:Fe molar ratio,
except that the most As was desorbed from the 1:0 Al:Fe hydroxide at each pH (Figure
72). Minimum As desorption was observed in the pH range of 5 to 9 with, increasing
desorption at both lower and higher pH values. Desorption of As increased slightly as
Al:Fe molar ratio was increased. The adsorption envelopes (Figure 52) indicated that
adsorption of As was approximately quantitative in the pH range of 3 to 6.5 at the
identical As:(Al+Fe) molar ratio as used in this experiment; therefore, surface sites were
available for quantitative As adsorption under this condition. Phosphate and As have
similar chemical characteristics. For example, H3PO4 and H3AsO4 have similar pKa
values: the pKa values of H3PO4 are 2.15, 7.20, and 12.35, and those of H3AsO4 are 2.20,
6.97, and 11.53. At pH >8, As desorption might have been influenced by electrostatic
repulsion as the negative charge character of both the hydroxide surface and the As
species were increasing.
The As desorption trend was relatively independent of Al:Fe molar ratio;
however, As desorption slightly increased as Al:Fe molar ratio was increased (Figure
73). The similar desorption trends might be attributable to the probability that in all cases
the As was likely adsorbed to surface Fe , according to the previous adsorption
isotherm study. The minimum desorption of As was observed at approximately pH 9.5,
which corresponds with the pH of maximum As I adsorption by Fe hydroxide (Figure 54;
also, Ferguson and Anderson, 1974, and Raven et al., 1998). Both the adsorption
V
V
V
V
V
V V
V
V
III
III
III III
III
II
127
pH
3 4 5 6 7 8 9 10 11
Ars
enic
des
orbe
d, %
0
20
40
60
80
100
0:1 Al:Fe2:8 Al:Fe5:5 Al:Fe1:0 Al:Fe
Figure 72. Desorption envelopes of AsV with sodium phosphate solution at 375 : 0.05 : 1 P:As:(Al+Fe) molar ratio, as affected by Al:Fe molar ratios.
128
igure 73. Desorption envelopes of AsIII with sodium phosphate solution at 375 : 0.05 : 1
pH
3 4 5 6 7 8 9 10 11
Ars
enic
des
orbe
d, %
0
20
40
60
80
100
0:1 Al:Fe2:8 Al:Fe5:5 Al:Fe
FP:As:(Al+Fe) molar ratio, as affected by Al:Fe molar ratios.
129
maximum and the desorption minimum of AsIII correspond approximately with the pKa1
of AsIII of 9.2. In general, AsV was desorbed more readily than AsIII above pH 7.5;
whereas, AsIII was desorbed more efficiently below pH 7.5 (Figures 74 –76).
130
pH
3 4 5 6 7 8 9 10 11
Ars
enic
des
orbe
d, %
0
20
40
60
80
100
AsV
AsIII
Figure 74. Desorption envelopes of AsV and AsIII with sodium phosphate solution at 375 : 0.05 : 1 P:As:Fe molar ratio with the 0:1 Al:Fe hydroxide.
131
pH
3 4 5 6 7 8 9 10 11
Ars
enic
des
orbe
d, %
0
20
40
60
80
100
AsV
AsIII
Figure 75. Desorption envelopes of AsV and AsIII with sodium phosphate solution at 375 : 0.05 : 1 P:As:(Fe+Al) molar ratio with the 2:8 Al:Fe hydroxide.
132
pH
3 4 5 6 7 8 9 10 11
Ars
enic
des
orbe
d, %
0
20
40
60
80
100
AsV
AsIII
Figure 76 . Desorption envelopes of AsV and AsIII with sodium phosphate solution at 375 : 0.05 : 1 P:As:(Fe+Al) molar ratio with the 5:5 Al:Fe hydroxide.
133
CONCLUSIONS
Mineralogy and Stability of the Hydroxides as Affected by Al:Fe Molar Ratio
Particle size, crystallinity, and morphology of aggregates are affected by Al:Fe
molar ratio. Data collected in XRD and TEM studies indicated that Al substitution in
ferrihydrite results in smaller particle size and less dense aggregates compared to pure
ferrihydrite. The maximum quantitative Al substitution in the poorly crystalline
ferrihydrite structure was up to approximately 20 %. The relative stability of the 2:8
Al:Fe hydroxide relative to pure ferrihydrite indicates that maximum stability of
ferrihydrite might be achieved when the structure is slightly relaxed because of the
smaller size of Al3+ than Fe3+. Once the Al content exceeded the maximum substitution
level, gibbsite and bayerite are formed. Location of Al in the ferrihydrite structure might
greatly affect the adsorption of As; therefore, further study of the local distribution of Al
is needed. In soils in which most of the ferrihydrite is Al substituted, Al in the structure
is likely a major factor contributing to ferrihydrite solubility and its transformation into
crystalline phases.
Adsorption of AsV and AsIII as Affected by Al:Fe Molar Ratio
Differences in adsorption of AsV and AsIII as affected by Al:Fe molar ratio of the
hydroxide were observed. When Al was completely substituted in the poorly crystalline
Fe hydroxide structure, the difference in AsV adsorption behavior compared to that of
pure ferrihydrite was negligible. Adsorption of AsV decreased as Al:Fe molar ratio
increased once the maximum Al substitution in ferrihydrite was achieved. Because
gibbsite and bayerite were detected in the Al:Fe hydroxide with more than 30 % Al, the
cause of decrease in AsV adsorption was the lower concentration of surface sites for
134
adsorption. Although higher surface area might be expected with 2:8 Al:Fe hydroxide
compared to pure ferrihydrite, adsorption of AsV with the 2:8 Al:Fe hydroxide did not
exceed that of pure ferrihydrite.
The adsorption maximum of AsIII decreased approximately 4 % with a 10 %
increase in Al content, even when Al was completely substituted in the structure of the
ferrihydrite. This phenomenon is indicative of different modes of bonding on AsV and
AsIII on Al:Fe hydroxides. There might be a difference in bonding strength of AsIII on Al
within an Al:Fe hydroxide compared to pure Al hydroxides, since reduction of AsIII
adsorption should have been proportional to the increase in Al content. A possible
heterogeneous distribution of Al in the structure might also influence the AsIII adsorption
behavior. A spectroscopic study would be useful to understand the local chemistry of
bonding. In addition, a better understanding of the distribution of Al within ferrihydrite
would help to elucidate the adsorption behavior of AsIII on Al:Fe hydroxides.
The retention of AsV at pH > 7 was significantly improved in the presence of the
counterion Ca2+ compared to Na+, probably due to the more rapid decay in repulsive
potential with distance from the surface with the former system. Counterion
concentration did not significantly affect AsV adsorption. The negligible influence of
counterion concentration is indicative of inner-sphere complexation.
Desorption of AsV and AsIII as Affected by Al:Fe Molar Ratio
Phosphate desorbed both AsV and AsIII from all Al:Fe hydroxides; however,
quantitative desorption was never obtained. The efficiencies of AsV and AsIII extraction
by phosphate were lowest at pH 5 to 9 and pH 9.5, respectively. In general, more AsV
was desorbed compared to AsIII above pH 7.5; whereas, AsIII was desorbed more
135
efficiently below pH 7.5. Desorption of both AsV and AsIII increased slightly with
increase in Al:Fe molar ratio, which indicates that As adsorbed on the Al portion of the
hydroxide might be more readily desorbed. The results of this study indicate that
phosphate significantly enhanced the release potential of both AsV and AsIII. In order to
understand As release in natural systems, desorption of As with other oxyanions such as
sulfate might to useful.
Implications to the Water Treatment
Effective water treatment requires the efficient removal of As and the secure
disposal of waste. Effective removal of AsV can be achieved using either ferrihydrite or
coprecipitated 2:8 Al:Fe hydroxide, as their adsorption behaviors were similar; however,
removal of AsIII is more effective with the pure Fe system. When AsIII is present,
oxidation of AsIII into AsV would be required to optimize the removal of As in the Al/Fe
system. The 2:8 Al:Fe molar ratio adsorbent is less soluble and more stable against
transformation into crystalline phases, which is advantageous for waste disposal.
In a reduced environment as might exist at a waste disposal site, both AsV and
FeIII are subject to be reduced into AsIII and FeII, respectively. Because FeII is soluble, Al
hydroxide will remain when an Al/Fe system is used; however, the weak affinity of AsIII
on Al hydroxide as shown in this study and other studies (Goldberg, 1986; Weerasooriya
et al., 2003) could contribute to As release. An increase of As release potential in a
reduced environment cannot be avoided whether pure ferrihydrite, pure Al hydroxide or
Al:Fe hydroxides are used.
The mineralogy of the hydroxides and adsorption/desorption behavior indicate the
possible utility of coprecipitated Al:Fe hydroxides in wastewater treatment. According to
136
the results of this study, 2:8 is a preferred Al:Fe molar ratio for As removal. Overall, the
chemistry of As with Al:Fe hydroxides is complex; therefore, further research is required.
It will be beneficial to study the retention of As that is coprecipitated with Al:Fe
hydroxides, to simulate the water treatment scenario. In addition, the relative impacts of
Al and Fe on the reduction of AsV should be more thoroughly investigated. The impact
of structural Al on the rate of FeIII reduction is not known. There is a possibility that Al
might slow the rate of the reduction of FeIII into FeII by retarding the ease of electron
transfer within the system, because Al is not affected by redox processes.
137
REFERENCES
Alam, M.G.M., S. Tokunaga, and T. Maekawa. 2001. Extraction of arsenic in a synthetic arsenic-contaminated soil using phosphate. Chemosphere 43:1035-1041.
Anawar, H.M., J. Akai, K.M.G. Mostofa, S. Safiullah, and S.M. Tareq. 2002. Arsenic poisoning in groundwater-health risk and geochemical sources in Bangladesh. Environ. Int. 27:597-604.
Anderson, M.A., J.F. Ferguson, and J. Gavis. 1976. Arsenate adsorption on amorphous aluminum hydroxide. J. Colloid Interface Sci. 54:391-399.
Arai, Y., E.J. Elzinga, and D.L. Sparks. 2001. X-ray absorption spectroscopic investigation of arsenite and arsenate adsorption at the aluminum oxide-water interface. J. Colloid Interface Sci. 235:80-88.
Bothe, V.J., and P.W. Brown. 1999. Arsenic immobilization by calcium arsenate formation. Environ. Sci. Technol. 33:3806-3811.
Bowden, J.W., A.M. Posner, and J.P. Quirk. 1977. Ionic adsorption on variable charge mineral surfaces. Theoretical-charge development and titration curves. Aust. J. Soil Res. 15:121-136.
Carrero, P., A. Malave, J.L. Burguera, M. Burguera, and C. Rondon. 2001. Determination of various arsenic species by flow injection hydride generation atomic absorption spectrometry: Investigation of the effects of the acid concentration of different reaction media on the generation of arsines. Anal. Chim. Acta 438:195-204.
Cheng, R.C., S. Liang, H.C. Wang, and M.D. Beuhler. 1994. Enhanced coagulation for arsenic removal. J. Am. Water Works Assoc. 86:79-90.
Cherry, J.A., A.U. Shaikh, D.E. Tallman, and R.V. Nicholson. 1979. Arsenic species as an indicator of redox conditions in groundwater. J. Hydrol. 43:373-392.
Colombo, C., and A. Violante. 1996. Effect of time and temperature on the chemical composition and crystallization of mixed iron and aluminum species. Clays Clay Miner. 44:113-120.
138
Cullen, W.R., and K.J. Reimer. 1989. Arsenic speciation in the environment. Chem. Rev. 89:713-764.
Davenport, J.R., and F.J. Peryea. 1991. Phosphate fertilizers influence leaching of lead and arsenic in a soil contaminated with lead arsenate. Water Air Soil Pollut. 57:101-110.
Dixit, S., and J.G. Hering. 2003. Comparison of arsenic(V) and arsenic(III) sorption onto iron oxide minerals: Implications for arsenic mobility. Environ. Sci. Technol. 37:4182-4189.
Edwards, M. 1994. Chemistry of arsenic removal during coagulation and Fe-Mn oxidation. J. Am. Water Works Assoc. 86:64-78.
Essington, M.E. 2004. Soil and water chemistry: An integrative approach. CRC Press, Boca Raton, FL.
Fendorf, S., M.J. Eick, P. Grossl, and D.L. Sparks. 1997. Arsenate and chromate retention mechanisms on goethite .1. Surface structure. Environ. Sci. Technol. 31:315-320.
Ferguson, J.F., and M.A. Anderson. 1974. Chemical forms of arsenic in water supplies and their removal, p. 137-158, In A. J. Rubin, (ed). Chemistry of water supply, treatment, and distribution. Ann Arbor, MI., Ann Arbor Science Publishers.
Fey, M.V., and J.B. Dixon. 1981. Synthesis and properties of poorly crystalline hydrated aluminous goethites. Clays Clay Miner. 29:91-100.
Francesconi, K., and D. Kuehnelt. 2002. Arsenic compounds in the environment, p. 51-94, In W. T. Frankenberger, (ed). Environmental chemistry of arsenic. Marcel Dekker, Inc., New York.
Goldberg, S. 1986. Chemical modeling of arsenate adsorption on aluminum and iron-oxide minerals, pp. 1154-1157 Soil Sci. Soc. Am. J., Vol. 50.
Goldberg, S., and C.T. Johnson. 2001. Mechanisms of arsenic adsorption on amorphous oxides evaluated using macroscopic measurements, vibrational spectroscopy, and surface complexation modeling. J. Colloid Interface Sci. 234:204-216.
139
Goyer, R.A., and T.W. Clarkson. 2001. Toxic effect of metals, p. 811-867, In C. D. Kalaassen, (ed). Casarett & Doull's Toxicology: The basic science of poisons, 6 ed. The McGraw-Hill Companies, Inc., New York.
Gulledge, J.H., and J.T. O'Connor. 1973. Removal of arsenic(V) from water by adsorption on aluminum and ferric hydroxides. J. Am. Water Works Assoc. 65:548-552.
Gupta, S.K., and K.Y. Chen. 1978. Arsenic removal by adsorption. J. Water Pollut. Control Fed. 50:493-506.
Hammer, M.J., and M.J.J. Hammer. 2001. Water and wastewater technology. 4 ed. Princeton-Hall, Inc., Upper Saddle River, NJ.
Harrison, J.B., and V.E. Berkheiser. 1982. Anion interactions with freshly prepared hydrous iron-oxides. Clays Clay Miner. 30:97-102.
Hering, J.G., P.Y. Chen, J.A. Wilkie, and M. Elimelech. 1997. Arsenic removal from drinking water during coagulation. J. Environ. Eng. 123:800-806.
Hering, J.G., P.Y. Chen, J.A. Wilkie, M. Elimelech, and S. Liang. 1996. Arsenic removal by ferric chloride. J. Am. Water Works Assoc. 88:155-167.
Hingston, F.J., A.M. Posner, and J.P. Quirk. 1971. Competitive adsorption of negatively charged ligands on oxide surfaces. Discuss. Faraday Soc. 52:334-342.
Hsia, T.H., S.L. Lo, and C.F. Lin. 1992. As(V) adsorption on amorphous iron oxide: triple layer modeling. Chemosphere 25:1825-1837.
Hsu, P.H. 1989. Aluminum hydroxides and oxyhydroxides, p. 331-378, In J. B. Dixon and S. B. Weed, (eds). Minerals in soil environments, 2 ed. SSSA, Madison, WI.
Huang, P.M., M.K. Wang, N. Kampf, and D.G. Schulze. 2002. Aluminum hydroxides, p. 261-289, In J. B. Dixon and D. G. Schulze, (eds). Soil mineralogy with environmental applications. SSSA, Madison, WI.
140
Inskeep, W.P., T.R. McDermott, and S. Fendorf. 2002. Arsenic (V)/(III) cycling in soils and natural water: Chemical and microbiological processes, p. 183-215, In W. T. Frankenberger, (ed). Environmental chemistry of arsenic. Marcel Dekker, New York.
Jackson, B.P., and W.P. Miller. 2000. Effectiveness of phosphate and hydroxide for desorption of arsenic and selenium species from iron oxides. Soil Sci. Soc. Am. J. 64:1616-1622.
Jain, A., and R.H. Loeppert. 2000. Effect of competing anions on the adsorption of arsenate and arsenite by ferrihydrite. J. Environ. Qual. 29:1422-1430.
Jain, A., K.P. Raven, and R.H. Loeppert. 1999. Response to comment on Arsenite and Arsenate adsorption on ferrihydrite: Surface charge reduction and net OH-release stochiometry. Environ. Sci. Technol. 33:3696-3696.
Jing, C.Y., G.P. Korfiatis, and X.G. Meng. 2003. Immobilization mechanisms of arsenate in iron hydroxide sludge stabilized with cement. Environ. Sci. Tchnol. 37:5050-5056.
Kosmas, C.S., D.P. Franzmeier, and D.G. Schulze. 1986. Relationship among derivative spectroscopy, color, crystallite dimensions, and Al substitution of synthetic goethites and hematites. Clays Clay Miner. 34:625-634.
Leist, M., R.J. Casey, and D. Caridi. 2000. The management of arsenic wastes: Problems and prospects. J. Hazardous Materials 76:125-138.
Lindsay, W.L. 1979. Chemical equilibria in soils. John Wiley & Sons, New York.
Liu, F., A. De Cristofaro, and A. Violante. 2001. Effect of pH, phosphate and oxalate on the adsorption/desorption of arsenate on/from goethite. Soil Sci. 166:197-208.
Loeppert, R.H., and W.P. Inskeep. 1996. Iron, In D. L. Sparks, (ed). Methods of soil analysis, part 3, chemical methods, Vol. 5, 3 ed. SSSA, Madison, WI.
Loeppert, R.H., A. Jain, M.A. Abd El-Haleem, and B.K. Biswas. 2002. Quantity and speciation of arsenic in soils by chemical extraction, In Y. Cai, (ed). Biogeochemistry of environmentally important elements. American Chemical Society, Washington, DC.
141
Manceau, A. 1995. The mechanism of anion adsorption on iron-oxides - evidence for the bonding of arsenate tetrahedra on free Fe(O,0H)(6) edges. Geochim. Cosmochim. Acta. 59:3647-3653.
Manning, B.A., and S. Goldberg. 1996. Modeling competitive adsorption of arsenate with phosphate and molybdate on oxide minerals. Soil Sci. Soc. Am. J. 60:121-131.
Manning, B.A., S.E. Fendorf, and S. Goldberg. 1998. Surface structures and stability of arsenic(III) on goethite: spectroscopic evidence for inner-sphere complexes. Environ. Sci. Technol. 32:2383-2388.
Masscheleyn, P.H., R.D. Delaune, and W.H. Patrick. 1991a. A hydride generation atomic absorption technique for arsenic speciation. J. Environ. Qual. 20:96-100.
Masscheleyn, P.H., R.D. Delaune, and W.H. Patrick. 1991b. Effect of redox potential and pH on arsenic speciation and solubility in a contaminated soil. Environ. Sci. Technol. 25:1414-1419.
McBride, B.C. 1997. A critique of diffuse double layer models applied to colloid and surface chemistry. Clays Clay Miner. 45:598-608.
McBride, M.B. 1994. Environmental chemistry of soils. Oxford University Press, New York.
Meng, X., G.P. Korfiatis, C. Jing, and C. Christodoulatos. 2001. Redox transformations of arsenic and iron in water treatment sludge during aging and TCLP extraction. Environ. Sci. Technol. 35:3476-3481.
Mott, C.J.B. 1981. Anion and ligand exchange, p. 179-219, In D. J. Greenland and M. H. B. Hayes, (eds). The chemistry of soil processes. John Wiley & Sons Ltd, Chichester, UK.
NRC. (National Research Council) 2001. Arsenic in drinking water National Academic Press, Washington, DC.
Nriagu, J.O. 2002. Arsenic poisoning through the ages, p. 1-26, In W. T. Frankenberger, (ed). Environmental chemistry of arsenic. Marcel Dekker, Inc., New York.
142
Paige, C.R., W.J. Snodgrass, R.V. Nicholson, and J.M. Scharer. 1996. The crystallization of arsenate-contaminated iron hydroxide solids at high pH. Water Environ. Res. 68:981-987.
Parks, J.L., J. Novak, M. Macphee, C. Itle, and M. Edwards. 2003. Effect of Ca on As release from ferric and alum residuals. J. Am. Water Works Assoc. 95:108-118.
Pierce, M.L., and C.B. Moore. 1982. Adsorption of arsenite and arsenate on amorphous iron hydroxide. Water Res. 16:1247-1253.
Raven, K.P., A. Jain, and R.H. Loeppert. 1998. Arsenite and arsenate adsorption on ferrihydrite: kinetics, equilibrium, and adsorption envelopes. Environ. Sci. Technol. 32:344-349.
Reed, B.E., and M.R. Matsumoto. 1993. Modeling cadmium adsorption by activated carbon using the Langmuir and Freundlich isotherm expressions. Separation science and technology. 28:2179-2195.
Sadiq, M. 1997. Arsenic chemistry in soils: An overview of thermodynamic predictions and field observations. Water Air Soil Pollut. 93:117-136.
Samanta, G., T.r. Chowdhury, B.K. Mandal, B.K. Biswas, U.K. Chowdhury, g.K. Basu, C.r. Chanda, D. Lodh, and D. Chakraborti. 1999. Flow injection hydride generation atomic absorption spectrometry for determination of arsenic in water and biological samples from arsenic-affected districts of West Bengal, India, and Bangladesh. Microchem. J. 62:174-191.
Schulze, D.G. 1984. The influence of aluminum on iron oxides. VIII. Unit-cell demensions of Al-substituted goethites and estimation of Al from them. Clays Clay Miner. 32:36-44.
Schulze, D.G., and U. Schwertmann. 1984. The influence of aluminum on iron oxides: X. Properties of Al-substituted goethite. Clay Miner. 19:521-539.
Schwertmann, S., and R.M. Taylor. 1989. Iron oxides, p. 379-438, In J. B. Dixon and S. B. Weed, (eds). Minerals in soil environment, 1 ed. SSSA, Madison, WI.
Schwertmann, U., and R.M. Cornell. 1991. Iron oxides in the laboratory: Preparation and characterization VCH, New York.
143
Schwertmann, U., and E. Murad. 1983. Effect of pH on the formation of goethite and hematite from ferrihydrite. Clays Clay Miner. 31:277-284.
Schwertmann, U., J. Friedl, H. Stanjek, and D.G. Schulze. 2000. The effect of Al on Fe oxides. XIX. Formation of Al-substituted hematite from ferrihydrite at 25oC and pH 4 to 7. Clays Clay Miner. 48:159-172.
Smedley, P.L., and D.G. Kinniburgh. 2002. A review of the source, behaviour and distribution of arsenic in natural waters. Appl. Geochem. 17:517-568.
Smith, A.H., P.A. Lopipero, M.N. Bates, and C.M. Steinmaus. 2002a. Public health - Arsenic epidemiology and drinking water standards. Science 296:2145-2146.
Smith, A.H., C. Hopenhayn-Rich, M.N. Bates, H.M. Goeden, I. Hertz-Picciotto, H.M. Duggan, R. Wood, M.J. Kosnett, and M.T. Smith. 1992. Cancer risks from arsenic in drinking water. Environ Health Persp. 97:259-267.
Smith, E., R. Naidu, and A.M. Alston. 2002b. Chemistry of inorganic arsenic in soils: II. Effect of phosphorous, sodium, and calcium on arsenic sorption. J. Environ. Qual. 31:557-563.
Sparks, D.L. 2003. Environmental soil chemistry. 2nd ed. Academic Press, San Diego, CA.
Sun, X.H., and H.E. Doner. 1996. An investigation of arsenate and arsenite bonding structure on goethite by FTIR. Soil Sci. 161:865-872.
Tokunaga, S., S. Yokoyama, and S.A. Wasay. 1999. Removal of Arsenic(III) and Arsenic(V) Ions from Aqueous Solutions with Lanthanum(III) Salt and Comparison with Aluminum(III), Calcium(II), and Iron(III) Salts. Water Environ. Res. 71:299-306.
USEPA. (United States Environmental Protection Agency) 2001. 40 CRF parts 9, 141, and 142, National Primary Drinking Water Regulations; Arsenic and Clarifications to Compliance and New Source Contaminant Monitoring; Proposed Rule. Fe Regist 66, Washington, DC.
Van Raij, B., and M. Peech. 1972. Electrochemical properties of some oxisols and alfisols of the tropics. Soil Sci. Soc. Am. Proc. 36:587-593.
144
Veith, J.A., and G. Sposito. 1977. On the use of the Langmuir equation in the interpretation of "adsorption" phenomena. Soil Sci. Soc. Am. J. 41:697-702.
Violante, A., P. Massimo, and R. Raffaella. 2002. Factors affecting arsenate adsorption/desorption on/from variable charge minerals and soils, Thailand.
Wagman, D.D., H.H. Evans, V.B. Parker, R.H. Schumm, I. Harlow, S.M. Bailey, K.L. Churney, and R.L.J. Butall. 1982. The NBS tables of chemical thermodynamic properties - selected values for inorganic and C-1 and C-2 organic-substances in SI units. J. Phys. Chem. Ref. Data 11:392.
Waychunas, G.A., B.A. Rea, C.C. Fuller, and J.A. Davis. 1993. Surface chemistry of ferrihydrite, Part l. EXAFS studies of the geometry of coprecipitated and adsorbed arsenate. Geochim. Cosmochim. Acta. 57:2251-69.
Weerasooriya, R., H.J. Tobschall, H.K.D.K. Wijesekara, E.K.I.A.U.K. Arachchige, and K.A.S. Pathiratne. 2003. On the mechanistic modeling of As(III) adsorption on gibbsite. Chemosphere 51:1001-1013.
Wefers, K., and G.B. Bell. 1972. Oxides and hydroxides of aluminum: Technical paper No. 19. Alcoa, East St. Louis, IL.
Woolson, E.A., J.H. Axley, and P.C. Kearne. 1973. Chemistry and phytotoxicity of arsenic in soils: Effects of time and phosphorous. Soil Sci. Soc. Am. J. 37:254-259.
Yang, C. 1998. Statistical mechanical study on the Freundlich isotherm equation. J. Colloid Interface Sci. 208:379-387.
145
VITA
Yoko Masue received her B.S. in plant and environmental soil sciences at Texas
A&M University in College Station, Texas in August 2002. After graduating, she
worked as a research assistant at Texas A&M University in pursuit of her M.S. degree in
soil science under the direction of Dr. R. H. Loeppert. During her study, she taught
undergraduate introductory soil science laboratories in 2003 under Dr. C.T. Hallmark.
Additionally, she has presented her thesis research at the 7th International Conference on
the Biogeochemistry of Trace Elements in Uppsala, Sweden, in June 2003, and the
American Society of Agronomy meeting in Denver, Colorado in November 2003. In
September 2004, she will begin work on her Ph.D. in geological and environmental
sciences at Stanford University in Stanford, California.
Permanent Address: 1-12-3 Kanaiwanishi, Kanazawa-shi, Ishikawa-ken, 920-0337, Japan